Eur. J. Biochem.

239, 356-361 (1996)

0 FEBS 1996

The hemolysin B protein, expressed in Saccharomyces cerevisiae,

accumulates in binding-protein(BiP)-containing structures
Ralf KOLLING and Cornelis P. HOLLENBERG

Institut fur Mikrobiologie, Heinrich-Heine-Universitat Dusseldorf, Germany

(Received 12 February/6 May 1996) - EJB 96 0194/1

The hemolysin B (HlyB) protein of Escherichia cnli, a member of the ABC-transporter family, was
expressed in Succharomyces cerevisiue and tested for its ability to complement a defect in the a-factor
transporter Ste6. We found that HlyB was not able to restore mating ability to a STEh deletion strain. The
HlyB protein did not co-fractionate with Ste6 on sucrose gradients, indicating that improper localization of
the HlyB protein could contribute to the lack of complementation. Immunofluorescence experiments
suggest that HlyB is localized to structures derived from the endoplasmic reticulum (ER). The HlyBexpressing cells revealed a perinuclear staining typical of ER-localized proteins and intensely staining
ring-like structures (HlyB-bodies). Double-label immunofluorescence experiments show that the HlyB
structures also contain the ER binding protein (BiP), the product of the kAR2 gene. The HlyB protein,
however, did not co-fractionate with another ER marker protein, the NADPH cytochroine c reductase.
The HIyB bodies could be derivatives of a novel compartment of the early secretory pathway which
contains BiP but not other resident ER proteins. In this case, HlyB could serve as a tool for the biochemical characterization of this compartment.

Keywords: ATP-binding cassette transporter; hemolysin B ; Ste6 ; binding-protein bodies, secretion.

The hemolysin B (HIyB) protein of Escherichiu coli is required in conjunction with hemolysin D (HlyD) and TolC [l]
for the secretion of the 107-kDa hemolysin A protein. HlyB is
a member of the ABC-transporter superfamily. ABC-transporters are characterized by a motif consisting of six putative transmembrane segments and a domain with an ATP-binding site,
hence the name ATP binding cassette transporter [2].HlyB and
other toxin-secreting bacterial transporters are unique among the
ABC-transporters in that they are able to transport large polypeptides across membranes [3]. The HlyB protein has striking
similarity to the mammalian Mdr proteins. By cluster analysis,
it has been shown that HlyB is more closely related to the human
P-glycoprotein (Mdrl) than to other E. coli ABC-transporters,
for instance to HisP or MalK [4]. Based on codon usage, it has
been suggested that E. coli has only recently acquired the hlyB
gene from a distantly related organism [ 5 ] . Considering these
findings, it appears possible that at least part of the hlyB gene is
of mammalian origin.
We were interested to determine whether the E. coli HlyB
protein could substitute for the yeast a-factor transporter Ste6
[6, 71. Ste6 is also closely related to the mammalian multi-drugresistance (Mdr) proteins. To some extent, the proteins are even
Correspondence to R. Kolling, Institut fur Mikrobiologie, Geb.
26.1 2.01, Heinrich-Heine-Universitat Dusseldorf, Universitatsstrasse 1
~

D-40225 Dusseldorf, Germany

Fax: +49 211 811 5370.
Abbreviations. BiP, binding protein; DAPI, 4,6-diamidino-2-phenylindole; DPAP A, dipeptidylaminopeptidase A ; HlyB, hemolysin B ; ER,
endoplasmic reticulum; mdr, niultidrug resistance.
Enzymes. a-n-Mannosidase (Amsl) (EC 3.2.1.24); / h u b u n i t F1
ATPase (Atp2), H transporting ATP synthase (EC 3.6.1.34) ; dipeptidylaminopeptidase A (EC 3.4.14); NADPH cytochrome c reductase (EC
1.6.2.4); plasma membrane ATPase (Pmal ), H transporting ATPase
(EC 3.6.1.34).

functionally interchangeable, as demonstrated by the ability of

mouse mdr3 to complement a STE6 defect [8]. We found, however, that HlyB was not able to restore mating ability to a STE6
deletion mutant. Our data indicate that mislocalization of the
HlyB protein could contribute to the lack of complementation.
Immunofluorescence experiments reveal brightly staining
HlyB containing structures in the yeast cell. The HlyB-containing structures are reminiscent of the recently described BiP bodies [9]. BiP (the product of the KAR2 gene), which is a member
of the Hsp70 protein family, is a soluble endoplasmic reticulum
(ER) protein [lo, 111. When secretion is blocked between ER
and Golgi, a relocation of BiP from a typical uniform ER distrbution to a punctate pattern is observed. These punctate structures were named BiP bodies [9]. It was concluded that the BiP
bodies correspond to subdomains of the ER. Here, we show that
the HlyB bodies, which form upon overproduction of HlyB, also
contain BiP. Our results suggest that HlyB is retained either in
an ER subcompartment or in an ER-Golgi intermediate compartment. The HlyB protein could serve as a marker for the biochemical characterization of this novel yeast compartment.

MATERIALS AND METHODS

Media and strains. The yeast strains were grown in rich
medium (YPD) containing 1 % Bacto Yeast extract (Difco), 2 %
Bacto Peptone (Difco) and 2% glucose, or in minimal medium
(SD) containing 0.7 % Yeast nitrogen base without amino acids
(Difco), 1% Casamino acids (Difco) and 2% glucose supplemented with uracil and tryptophan as required. The genotypes
of the strains used are: DBY2063 MATa leu2 uru3, JPY201
MATa his3 leu2 lys2 ste6: :HIS3 trpl ura3.
Preparation of affinity purified anti-HlyB serum. The
preparation of TrpE-HlyB fusion protein from strain HBlOl

Kolling and Hollenberg ( E m J. Biochem. 239)

Fig. 1. Schematic representation of the used constructs. The basic

ABC-transporter unit is represented by long bars, the Ste6 spacer region
and the N-terminal extension of HlyB by short bars. HA = hemagglutinin tag, PDCl = PDCl promoter fragment. The detailed structure of
the constructs is described in Materials and Methods.

containing the plasmid pRKl34 was essentially as described for

TrpE-Mlpl protein [I 21.
Plasmid constructions. Several STE6-hlyB hybrids were
constructed. The sequence positions are taken from [6] for STE6
and from [13] for hlyB. To facilitate the construction of hybrids,
additional BamHI sites were introduced at the end of repeat 1
of STE6 (5- ...AATGAC2289GGATCC-3), at the beginning of
repeat 2 of STE6 (5-GGATCCTAATC...-3) (original STE6 or
hlyB sequences in bold print; sequences introduced during construction in normal print) and at the beginning of the HlyB repeat (5-GGATCCTG A4930TTGAA...-3) using the PCR technique. PCR fragments with one BamHI and one BglII end were
generated coding for the STE6 spacer segment (5-GGATCCATAATGT2293CTGAT ...TACAAAAA2612GATCT-3). These
different building blocks were combined in several ways
(Fig. 1). The plasmids pRK87 and pRK102 are based on the 2pvector YEp420 [14]. The recombinant protein from pRK87 is
expressed from the PDCl promoter. A TrpE-HlyB fusion was
constructed (pRK134) by inserting the 1.1-kb EcoRV-Hind111
fragment encoding the C-terminal part of HlyB into pATH2 [15]
cut with Hind111 and SmaI. The plasmid pRK168, based on the
2p-plasmid YEp420, contains the E. coli hlyB gene (positions
4427-6727) [I31 under the control of the PDCl promoter.
pRKl66 is identical to pRKl68 except for the presence of the
hemagglutinin tag (coding for the amino acids YPYDVPDYAAFL) inserted after the last codon of hlyB.
Immunofluorescence experiments [16]. Cells were grown
to exponential phase (A,,, = 0.5-0.8, 3-4X107/ml) in YPD or
SD/Casamino acids medium and fixed directly for 2-3 h with
formaldehyde (final concentration 5 %). The fixed cells were
washed once in spheroplasting buffer (1.O M sorbitol, SO mM
Hepes, 10 mM NaN,, pH 7 3 , resuspended in 1 ml spheroplasting buffer and spheroplasted for about 30 min with zymolyase (0.1 mg/ml, 100 U/mg), glusulase (30 pl/ml) and 10 mM dithiothreitol at 37C. The spheroplasts were extracted with 0.1 %
Triton X-100 for 5 min, then attached to a multiwell slide treated
with 0.1% polylysine (Sigma). The cells were first incubated
with the anti-hemagglutin mouse monoclonal primary antibody
[SCP-12CAS-I, Berkeley Antibody Co, Inc.; 1 :200 dilution in
10 mM NaH,PO,, 150 mM NaCl, pH 7.2 (NaCUP,) 1 mg/ml
BSA] for 90min, then another 90 min with fluoresceinisothiocyanate-conjugated anti-mouse secondary antibodies (Dianova, 1: 300 dilution in NaCI/P, 1 mg/ml BSA). Finally, the
cells were incubated for 5 min with 4,6-diamidino-2-phenylindole (DAPI) (1 pg/ml). For double-label immunofluorescence,

357

the cells were first incubated with a mixture of mouse anti-hemagglutin antibodies (1 :200) and rabbit anti-Kar2 antibodies
(1 :500). then with a mixture of anti-mouse IgG fluoresceinisothi0c:yanate-conjugated antibodies (1 : 300) and anti-rabbit
IgG rhcldamine-conjugated antibodies (1 : 300). The anti-Kar2
sera were a gift from Mark Rose (Princeton University,
Princeton, USA). Between the different steps, the cells were
washed extensively with NaCIP, using an aspirator. The cells
were mounted with mounting medium (1 mg/ml p-phenylendiamine, O.lXNaCl/P,, 90% glycerol, pH 9). The cells were examined with a Zeiss standard microscope equipped for epifluorescence and photographed with Kodak TMAX 400 black and
white film.
Fractionation experiments. Exponentially growing cells
(A,,,,, = 0.5-1.0, 3-6X107/ml) from a 100-ml culture were harvested, washed once in cold 10 mM NaN, and once in STEDIO
(lo%, by mass, sucrose, 10 mM Tris/CI, pH 7.6, 10 mM EDTA,
1 mM clithiothreitol). After resuspending the cells i n 0.5 ml
STED10 (+protease inhibitors; 0.5 pg/ml each of aprotinin,
antipain. chymostatin, leupeptin, pepstatin A and 1.6 pg/ml
benzamidine, 1 pg/ml phenanthroline and 170 pg/ml phenylmethylsiilfonylfluoride), glass beads were added up to the top of
the liquid and the suspension was vortexed for 2 min. After the
addition of another 1 ml STEDIO, the extract was transferred to
a fresh tube and spun at 500 g for 5 min to remove cell debris.
About 1 ml cleared cell extract was loaded onto a sucrose gradient prep,xed the following way. In a SW40 centrifuge tube, 4 ml
STEDSO (53% sucrose), STED35 (35% sucrose) and STED20
(20% sucrose) were layered on top of each other (STED =
sucrose, 10 mM Tris/Cl, pH 7.6, 1 mM EDTA, 1 mM dithiothreitol). The centrifuge tubes were slowly turned to a horizontal
position. After 3 -5 h of horizontal diffusion, the gradients were
again turned into an upright position, loaded with extract and
spun for 13-17 h at 30000 rpm in a SW40 rotor at 4C. 700p1 fractions were collected from the top of the gradient.
Enzyme assays. NADPH cytochrome c reductase. The redox-change in the absorbance spectrum of cytochrome c upon
reduction by the enzyme was measured. The reaction mixture
consisted of 100 mM KP,, pH 7.0, 1 mM KCN, NADPH
(100 pM) and cytochrome c (25 pM) in its oxidized form. 2050 p1 gradient fractions are added and the initial rate of increase
in absorlxnce at 550 nm was registered in a spectrophotometer.
a-Mannosiduse. The assay measures the generation of the
yellow p-nitrophenol from the non-coloured glycoside. To
2.50 pl of the reaction mixture (SO mM KP,, 0.04% Triton X100, 1 nnM f7-nitrophenyl-a-mannoside), 50 p1 of the gradient
fractions were added. After incubating the samples at 30C for
2 h, 600 pl 2% Na,CO, was added to stop the reaction. Then,
the absorbance at 400 nm was determined.
Mating assay. Equal numbers of exponentially growing
MATa arid MATa cells (1 X 107)were mixed. The cells were pelleted, overlayed with 5 ml YPD medium and incubated for 4 h
at 30C. Aliquots of serial dilutions were plated on rich medium
to deterrnine the total cell number and on selective plates allowing only for growth of the zygotes. The fraction of cells which
were able to grow on the selective plates was determined. The
frequency of zygote formation for a wild-type MATa mating
partner was usually about 0.12. For a Aste6 MATa strain no zygotes were detected among 5 X 10, cells (frequency < 2X 1O-7).

RESULTS
Expression of hemolysin B in yeast. The hemolysin B gene of
E. coli was cloned behind the PDCl promoter on a multi-copy
plasmid (pRK168). The upstream ATGs 5 of the start codon

358

Kolling and Hollenberg ( E M J. Biochenz. 239)

Fig. 2. Expression of HlyB and HlyB/Ste6 fusions. The strain JPY201

was transformed with the plasmids indicated. Cell extracts were prepared
and analyzed by Western blotting (A) with anti-hemolysin B serum or
(B) with anti-hemagglutinin tag serum.

were removed by exonuclease Ba131 digestion to prevent improper tranvlation initation. To demonstrate HlyB expression in
yeast, antibodies were raised against a TrpE-HlyB fusion protein. The antibodies were affinity purified against fusion protein
bound to a column. With these antibodies, a specific HlyB signal
could be detected on Western blots with whole cell extracts
(Fig. 2A, lane 1). The signal, however, was only weak. To improve the detection of the hemolysin B protein, we constructed
an epitope-tagged variant of hemolysin B by inserting the hemagglutinin-tag at the C-teiminus of HlyB (pRK166). The HlyBhemagglutinin protein could be detected in whole cell extracts
by Western blotting with anti-(hemolysin B) (Fig. 2A, lane 2)
and with anti-hemagglutinin-tag antibodies (Fig. 2 B, lane 6).
The addition of the hemagglutinin-tag seems to somehow stabilize the protein, since we observed a higher protein level for the
HlyB-hemagglutinin variant compared to normal HlyB protein
(Fig. 2A, lanes 1 and 2). The hemagglutinin-tagged variant
shows a slightly slower mobility on SDS/polyacrylamide gels,
due to the presence of 14 additional hemagglutinin-tag amino
acids.

HlyB expression does not promote mating. Sequence comparisons suggested a close relationship between hemolysin B and
Ste6 [17]. Hydropathy profiles of Ste6 and HlyB are almost superimposible, indicating that the structures of the proteins are
very similar. Quantitative mating experiments with HlyB
(JPY201 dste6/pRK168) or HlyB-HA (JPY201 dste6/166) expressing strains, however, demonstrated that HlyB is not able to
replace the Ste6 protein in a-factor secretion. The same result
was obtained with a chimeric protein (pRK87, Figs 1 and 2A,
lane 4) where the 150-amino-acid N-terminal extension of HlyB
had been replaced by the corresponding 100-amino-acid segment of Ste6 connecting the two ABC-transporter repeats.
One possible reason for the lack of mating activity could be
that the HlyB protein is mislocalized to a cellular compartment
incompatible with a-factor secretion. In an earlier report, it was
shown that a functional transporter can be reconstituted by expressing the two halves of Ste6 separately on two different polypeptides [IS]. Both halves are, however, required for
a-factor secretion. Obviously, both halves of Ste6 carry the
targeting information for their correct cellular localization,
otherwise independent expression of the two half-molecules
would not have led to the reconstitution of a functional transporter. To make use of the targeting information contained in the
Ste6 repeats, a chimeric protein was constructed where the Cterminal Ste6 repeat was replaced by the corresponding HlyB
sequences (Fig. 1). As the other constructs, this chimeric protein
(pRK102) was expressed (Fig. 2A, lane 3) but was non-functional.

6 0 ~

*.----30
20

10

10

12

14

16

18

Fractions
Fig. 3. Fractionation of membranes on a sucrose gradient. A whole
cell extract of strain DBY2063, transformed with the plasmid pRK166,
was fractionated on a sucrose gradient. The gradient fractions were analyzed by Western blotting with antibodies against (A) the hemagglutinin
tag, (B) the p-subunit of F1 ATPase (Atp2), (C) the Ste6 protein, (D)
dipeptidyl aininopeptidase A (DPAP A) and (E) plasma membrane

ATPase (Pmal). (F) Enzyme activity of NADPH cytochrome c reductase

(0)and of a-mannosidase ( 0 )in arbitrary units and the density of the
gradient fractions (+; sucrose, by mass).

Intracellular localization of HIyB. The intracellular distribution of the HlyB protein was examined by cell fractionation experiments. Whole cell extracts were fractionated on sucrose gradients. The gradient fractions were analyzed for the presence of
different marker proteins by Western blotting (Fig. 3, A-E) or
by enzyme assay (Fig. 3 F). The peak of the HlyB-hemagglutinin
distribution was found in fractions 9 and 10 (31-33% sucrose).
HlyB, therefore, does not co-localize with the endogenous ABCtransporter Ste6 which has a distribution with the main peak in
fractions 7 and 8 (27-29% sucrose). As demonstrated previously 1191, the Ste6 protein co-fractionates with the Golgi
marker protein dipeptidyl aminopeptidase A (DPAP A) [20].
Also, the distribution of HlyB-hemagglutinin on the gradient is
clearly different from the distribution of the ER marker NADPH
cytochrome c reductase [21] (peak in fractions 3 and 4, 2022 % sucrose), the plasma membrane marker plasma membrane
ATPase (Pmal) 1221 (peak in fractions 15-18, 45-51%
sucrose) or the vacuolar marker protein a-mannosidase (Ams 1 )

Kolling and Hollenberg (ELK J . B,'oclzem 239)

HlyB-HA

DAPl

359

cating that a certain fraction of the HlyB-hemagglutinin protein

was associated with the ER. The ring-like HlyB-hemagglutinincontaining structures are quite unusual and do not resemble any
of the well characterized yeast organelles. They are reminiscent
of the recently described BiP bodies [9]. To see whether the
HlyB bodies represent structures analogous to the BiP bodies,
we investigated by double-label immunof-luorescence whether
HlyB bodies contain BiP (Kar2). We found a perfect co-localization of HlyB-HA and Kar2 staining (Fig. 5). The presence of
BiP in the HlyB bodies suggests that these structures are derived
from the ER.

DISCW3SION
Fig. 4. Irnrnunofluorescence staining of HlyB-HA. Cells of strain
DBY2063 transformed with the plasmid pRK166 were stained with antihemagglutinin tag primary antibodies and fluorescein-isothiocyanateconjugated anti-rnouse secondary antibodies. No staining was observed
with ?train DBY2063 expressing the pRK168 hemolysin B variant without hemagglutinin tag (not shown). The nuclear DNA was stained with
DAPI.

HlyB-HA

KarP

DAPI

Fig. 5. Double-label immunofluorescence staining of HlyB-HA and

KarZ. Cells of strain DBY2063 transformed with plasmid pRK166 were
stained with antibodies against the hemagglutinin tag and against Kar2.
The mouse hemagglutinin-tag antibodies were detected with anti-mouse
IgG fluorescein-isothiocyanate-conjugated secondary antibodies using a
fluorescein isothiocyanate filter set; the rabbit Kar2 antibodies were detected with anti-rabbit IgG rhodamine-conjugated secondary antibodies
using a rhodamine filter set. The nuclear DNA was stained with DAPI.

1231 (peak in fractions 7 and 8, 27-29% sucrose). The marker

protein closest to the HlyB-hemagglutinin protein was the p subunit of the mitochondria1 F1 ATPase (Atp2). The profiles of the
gradient distribution of the two proteins, however, was different.
We, therefore, conclude that the HlyB-hemagglutinin protein
does not co-fractionate with any of the marker proteins tested.
This may indicate that HlyB-hemagglutinin is localized to a
novel intracellular compartment.
Next, we examined the localization of the HlyB-hemagglutinin protein by immunofluorescence microscopy. Since the HlyBhemagglutinin protein was expressed from a 2p-plasmid, which
shows a marked fluctuation in copy number, the intensity of
staining varied from cell to cell. In about 5-10% of the cells,
a small number of intensely staining ring-like structures were
observed (in most cases about 1-6kell; Fig. 4). In addition to
the ring-like structures, we observed a perinuclear staining, indi-

There is a substantial degree of sequence identity between

HlyB and the yeast a-factor transporter Ste6 [17]. We therefore
tested whether HlyB could replace Ste6 i n a-factor secretion.
HlyB, alone or i n combination with the N-terminal Ste6 repeat,
was, however, completely inactive in the mating assay, indicating that HlyB, in contrast to mammalian mdr3 (81, is not able
to replace Ste6. One reason for this lack of complementation
could be improper intracellular localization of the HlyB protein.
We found that HlyB and Ste6 do not co-localize on sucrose gradients irtdicating that the two proteins are localized to different
cellular compartments.
At present, it is unclear where Ste6 performs its function.
Previous results have demonstrated that Ste6 is mainly localized
to an intracellular compartment, presumably the Golgi complex
[19]. However, there is also evidence that Ste6 has access to the
plasma inembrane. Both, the intracellular compartment and the
plasma inembrane could be the site of Ste6-dependent a-factor
translocation. The HlyB protein appears to be localized to a different compartment since HlyB neither co-fractionates with the
Golgi marker DPAP A nor with the plasma membrane marker
Pmal. The localization in this other compartment could be incompatible with a-factor secretion.
Our immunofluorescence experiments suggest that the HlyB
compartment is derived from the ER. Most of the HlyB-expressing cell!; show a perinuclear staining, the typical pattern observed fix an ER protein [ I l l . This also indicates that the HlyB
protein, which does not have a N-terminal hydrophobic signal
sequence, is inserted into the ER membrane. In addition to the
perinucll-ar staining, we observe in some of the cells brightly
staining ring-like structures. These structures also contain the
Kar2 protein (or BiP). Kar2 is normally retained in the ER
through its C-terminal HDEL retention signal [24]. Like other
HDEL containing proteins, Kar2 is thought to cycle between the
ER and an early cis-Golgi compartment [25]. Normally, Kar2
does not seem to progress beyond the cis-Golgi. The presence
of Kar2 in the HlyB bodies can, therefore, be taken as an indication that the HlyB-containing structures are either part of the
ER or correspond to an ER-Golgi intermediate compartment.
Interestingly, the HlyB protein does not co-fractionate with the
ER marker enzyme NADPH cytochrome c reductase. This may
indicate that specialized ER subdomains exist in yeast. However,
HlyB could also be localized to an intermediate compartment
between ER and Golgi. Such an ER-Golgi intermediate compartment (EIRGIC) has been well characterized morphologically and
biochemically in higher cells [26]. In yeast, the existence of an
intermediate compartment has not been demonstrated. However,
evidence has been presented that at least some resident ER proteins cycle through an early cis-Golgi compartment 1251. Given
the high degree of conservation of the secretory pathway between yl-ast and higher eukaryotic cells, it is reasonable to assume th,it yeast also has an ERGIC-like compartment. If HlyB

360

Kijlling and Hollenberg (Eur: J. Biochem. 239)

6. Kuchler, K., Sterne, R. E. & Thorner, J. (1989) Saccharomyes cereresides in such a compartment, it could be used as a marker for
visiae STE6 gene product: a novel pathway for protein export in
its biochemical characterization.
eukaryotic cells, EMBO J. 8, 3973-3984.
The brightly staining HlyB containing structures are reminis7. McGrath, J. P. & Varshavsky, A. (1989) The yeast STE6 gene encent of the recently described BiP bodies [9]. These structures,
codes a homologue of the mammalian multidrug resistance Pstained by anti-BiP (Kar2). were observed on two occasions,
glycoprotein, Nature 340, 400-404.
upon overproduction of the Secl2 protein and upon imposition
8 . Raymond, M., Gros, P., Whiteway, M. & Thomas, D. Y. (1992)
of a secretory block between the ER and Golgi in sec mutants.
Functional complementation of yeast ste6 by a mammalian multidrug resistance mdr gene, Science 256,232-233.
Usually, a relatively large number of small dots were observed.
9. Nishikawa. S., Hirata, A. & Nakano, A. (1994) Inhibition of endoHowever, there was one sec mutant (sec20) which had a pattern
plasmic reticulum (ER)-to-Golgi transport induces relocalization
that closely resembled our HlyB staining pattern. So, there is
of binding protein (BiP) within the ER to form the BiP bodies,
some justification for the view that HlyB bodies and BiP bodies
Molec. Biol. Cell 5, 1129-1143.
are analogous structures. Since BiP bodies are formed when
10. Rose, M. D., Misra, L. M. & Vogel, J. P. (1989) KAR2, a karyogamy
secretion between ER and Golgi is blocked, overproduction of
gene, is the yeast homolog of the mammalian BiPlGRP78 gene,
HlyB may similarly lead to a block early in the secretory pathCell 57, 1212-1221.
way. Since only few cells in the cell population exhibit HlyB 11. Preuss, D., Mulholland, J., Kaiser, C. A., Orlean, P., Albright, C.,
bodies, which presumably are the result of extreme HlyB overRose, M. D., Robbins, P. & Botstein, D. (1991 j Structure of the
yeast endoplasmic reticulum: localization of ER proteins using
production, it is, however, difficult to prove this assumption.
immunofluorescence and immunoelectron microscopy, Yeast 7,
Why is the HlyB protein retained early in the secretory path891 -91 1.
way ? According to the bulk-flow hypothesis, one would assume
12.
Kiilling,
R., Nguyen, T., Chen, E. Y. & Botstein, D. (1993) A new
that HlyB is transported to the default compartment, unless it
yeast gene with a myosin-like heptad repeat structure, Mol. Gen.
carries specific localization signals. For membrane proteins in
Genet. 237, 359-369.
yeast, the default compartment seems to be the vacuole [20]. 13. Hess, J., Wels, W., Vogel, M. & Goebel. W. (1986) Nucleotide seSince HlyB does not co-fractionate with the vacuolar marker uquence of a plasmid-encoded hemolysin determinant and its commannosidase, there is no indication for a vacuolar localization
parison with a corresponding chromosomal hemlysin sequence,
of HlyB. HlyB could be retained early in the secretory pathway,
FEMS Microbid. Lett. 34, 1 - 11.
due to the presence of an ER retention signal [27, 281. However, 14. Ma, H., Kunes, S., Schatz, P. J. & Botstein, D. (1987) Plasinid construction by homologous recombination in yeast, Gene (Amst.)
HlyB neither contains a C-terminal HDEL sequence nor a dily58, 201 -216.
sine motif. However, we cannot rule out the possibility that unknown retention signals exist in the HlyB protein. There are 15. Dieckmann, C. L. & Tzagoloff, A. (1985) Assembly of the mitochondrial membrane system. CBP6, a yeast nuclear gene necesindications that other ER retention signals exist in addition to
sary for the synthesis of cytochrome b, J . Biol. Chem. 260, 1513the known signals [29]. Retention of HlyB could also occur by
1520.
formation of oligomers. This retention mechanism has been pro- 16. Pringle, J. R., Preston, R. A., Adams, A. E. M., Stearns, T., Drubin,
posed for p63, a protein of the intermediate compartment [30]
D. G., Haarer, B. K. & Jones, E. W. (1989) Fluorescence microand for Golgi proteins [30, 311. Another possibility is that the
scopy methods for yeast, Methods Cell Biol. 31, 357-434.
HlyB protein is retained in the ER by components of the ER 17. Kuchler, K. & Thomer, J. (1992) Secretion of peptides and proteins
lacking hydrophobic signal sequences: the role of adenosine
quality control system, such as BiP or calnexin [32], because
tripliosphate-driven membrane translocators, Endocrine Rev. 13,
of incomplete folding or rnisfolding.
499 - 5 14.
In summary, we have shown that HlyB accumulates in a
novel structure of the early secretory pathway which could be 18. Berkower, C. & Michaelis, S. (1991) Mutational analysis of the
yeast a-factor transporter STE6, a member of the ATP binding
part of the ER or which could correspond to an ER-Golgi intercassette (ABC) protein family, EMBO J. 10, 3777-3785.
mediate compartment. HlyB could serve as a tool for the bio19. Kiilling, R. & Hollenberg, C. P. (1994) The ABC-transporter Ste6
chemical characterization of this compartment.
accumulates in the plasma membrane in a ubiquitinated form in
We thank Kevin Baker for the Atp2 antibodies, Trinad Chakraborty
for the hemolysin plasmids, Mark Rose for the Kar2 antibodies, Ramon
Serrano for the Pmal antibodies and for communicating his cell fractionation protocol and Tom Stevens for the DPAP A antibodies. This work
was supported in part by the BMBF project Stoffurnwandlungen mit Bioktrtalyscitoren and by Deutsche For.s-sc.hunRsRemeinschuft grant KO 9631
2-1 to R. K.

REFERENCES
1 . Wandersman, C. & Delepelaire, P. (1990) TolC, an Escherichia coli
outer membrane protein required for hemolysin secretion, Proc.
Natl Acad. Sci. USA 87, 4776-4780.
2. Higgins. C. F. (1992) ABC transporters: from microorganisms to
man. Annu. Rev. Cell Biol. 8, 67- 113.
3. Blight, M. A. & Holland, I. B. (1990) Structure and function of
haemolysin B, P-plycoprotein and other members of a novel family of membrane ~ranalocators,Mol. Microhiol. 4, 873 -880.
4. Gerlach. J . H., Endicott. J. A,. Juranka, P. F., Henderson, G., Sarangi, F., Deuchars. K. L. & Ling, V. (1986) Homology between
P-glycoprotein and a bacterial hnemolysin transport protein SUEgests a model for multidrug resistance, Nature 324, 485 -489.
5 . Felmlee, T., Pellett, S. & Welch, R. A . (1985) Nucleotide Sequence
of an Esc,herichia coli chromosomal hemolysin, .I. Bucteriol. 163,
94- 105.

endocytosis mutants, EMBO J. 13, 3261 -3271.

20. Roberts, C. J., Nothwehr, S. F. & Stevens, T. H. (1992) Membrane
protein sorting in the yeast secretory pathway: evidence that the
vacuole may be the default compartment, J. Cell Biol. 119, 6983.
21. Yabusaki, Y.,
Murakami, H. & Ohkawa, H. (1988) Primary structure
of Succharomyces cerevisiae NADPH-cytochrome P450 reductase deduced from nucleotide sequence of its cloned gene. J. Biochen7. (Tokyo) /03, 1004-1010.
22. Serrano, R., Kielland-Brandt, M. C . & Fink, G. R. (1986) Yeast
plasma membrane ATPase is essential for growth and has homology with (Na + K+), K - and Ca-ATPases, Nature 319, 689693.
23. Yoshihisa, T. & Anraku, Y. (1990) A novel pathway of import of amannosidase, a marker enzyme of vacuolar membrane, in
Sacchurotnyces cerevisiue, J . Biol. Chem. 265, 22 41 8 -22 424.
24. Hardwick, K. G., Lewis, M. J., Semenza. J., Neta, D. & Pelham, H.
R. B. (1990) ERDJ, a yeast gene required for the retention of
luminal endoplasmic reticulum proteins, affects glycoprotein processing in the Golgi apparatus, EMBO J. 9, 623-630.
25. Dean, N. & Pelham, H. R. B. (1990) Recycling of Proteins from the
Golgi compartment to the ER in yeast, J. Cell B i d . 111. 369377.
26. Lippincott-Schwartz, J. (1993) Bidirectional membrane traffic between the endoplasmic reticulum and Golgi apparatus, Trends
Cell B i d . 3. 81 -88.

Kolling and Hollenberg (Eur: J. Blochem. 239)

27. Munro, S. & Pelham, H. R. B. (1987) A C-terminal signal prevents
secretion of luminal ER proteins, Cell 48, 899-907.
28. Jackson, M. R., Nilsson, T. & Peterson, P. A. (1990) Identification
of a consensus motif for retention of transmembrane proteins in
the endoplasmic reticulum, EMBO J. 9, 3153-3162.
29. Parlati, F., Dominguez, M., Bergeron, J. J. M. & Thomas, D. Y.
(1995) Succharomyces cerevisiue CNEl encodes an endoplasmic
reticulum (ER) membrane protein with sequence similarity to calnexin and calreticulin and functions as a constituent of the ER
quality control apparatus, J. Biol. Chem. 270, 244-253.

361

30. Schweizer, A,, Rohrer, J., Hauri, H.-P. & Kornfeld, S. (1994) Retention of p63 in an ER-Golgi intermediate compartment depends on
the presence of all three of its domains and on its ability to form
oligomers, J. Cell Biol. 126, 25-39.
31. Weisz, 0. W., Swift, A. M. & Machamer, C. E. (1993) Oligomerization of a membrane protein correlates with its retention in the
Golgi complex, J. Cell Biol. 122, 1185- 1196.
32. Bergzron, J. J. M., Brenner, M. B., Thomas, D. Y. & Williams, D.
B. (1994) Calnexin: a membrane-bound chaperone of the endoplasmic reticulum, Trends Biol. Sci. 19, 124-128.