The EMBO Journal Vol.16 No.9 pp.2251–2261, 1997
The linker region of the ABC-transporter Ste6
mediates ubiquitination and fast turnover of the
at least a certain fraction of Ste6 travels to the plasma
membrane. This evidence is derived from experiments
with endocytosis mutants showing an accumulation of
Ste6 at the plasma membrane upon block of endocytosis.
Following endocytosis, the Ste6 protein seems to be
degraded in the vacuole (Berkower et al., 1994; Ko ¨lling
and Hollenberg, 1994).
Curiously, the Ste6 protein which accumulates at the
plasma membrane in endocytosis mutants is ubiquitinated.
What is the role of this ubiquitination? Most thoroughly
characterized is the role of ubiquitination in protein
turnover. Ubiquitin, a 76 amino acid polypeptide, is
covalently attached to protein substrates, thereby marking
the proteins for degradation by the 26S proteasome, an
(Ciechanover, 1994; Hochstrasser, 1995; Jentsch and
Schlenker, 1995). Despite extensive knowledge about the
enzymology of ubiquitin-dependent proteolysis, only a
limited number of natural substrates are known to date.
In yeast, the ubiquitin system has been implicated in the
degradation of several soluble proteins, including the
transcriptional regulators Matα2 (Chen et al., 1993) and
Gcn4 (Kornitzer et al., 1994), several yeast cyclins
(Deshaies et al., 1995; Seufert et al., 1995; Yaglom et al.,
1995), fructose 1,6-bisphosphatase (Schork et al., 1995),
the Gα subunit Gpa1 (Madura and Varshavsky, 1994) and
the GTP exchange factor Cdc25 (Kaplon and Jacquet,
1995). Degradation via the ubiquitin–proteasome pathway,
however, does not seem to be restricted to soluble proteins.
There are also a few examples of integral membrane
proteins that seem to be degraded via the proteasome.
One example that is of special interest in this context is
the human CFTR protein, because it is itself a member
of the ABC-transporter family. The CFTR protein seems
to have problems in attaining a properly folded conforma-
tion. A large fraction of the wild-type precursor protein
and virtually all of a mutant form of CFTR (∆F508),
found in the majority of cystic fibrosis patients, is retained
in the endoplasmic reticulum (ER) by the quality control
system and is apparantly degraded by the ubiquitin–
proteasome pathway (Jensen et al., 1995; Ward et al.,
1995). Similarly, a mutant form of the multispanning
membrane protein Sec61, a yeast protein involved in
protein translocation, seems to be degraded at the level
of the ER by the ubiquitin–proteasome pathway (Biederer
et al., 1996).
However, the consequence of ubiquitination is not
always the immediate degradation of the target proteins.
The immunoglobulin E receptor is ubiquitinated in
response to antigen binding and is deubiquitinated rapidly
upon receptor disengagement (Paolini and Kinet, 1993).
There is no evidence for a selective degradation of
the ubiquitinated receptors. The cycle of ubiquitination–
deubiquitination, therefore, seems to serve a purely regu-
Ralf Ko ¨lling1and Sascha Losko
Institut fu ¨r Mikrobiologie, Heinrich-Heine-Universita ¨t Du ¨sseldorf,
D-40225 Du ¨sseldorf, Germany
Upon block of endocytosis, the a-factor transporter
Ste6 accumulates in a ubiquitinated form at the plasma
membrane. Here we show that the linker region, which
connects the two homologous halves of Ste6, contains
a signal which mediates ubiquitination and fast turn-
over of Ste6. This signal was also functional in the
context of another plasma membrane protein. Deletion
of an acidic stretch in the linker region (‘A-box’)
strongly stabilized Ste6. The A-box contains a sequence
motif (‘DAKTI’) which resembles the putative endo-
cytosis signal of the α-factor receptor Ste2 (‘DAKSS’).
Deletion of the DAKTI sequence also stabilized Ste6
but, however, not as strongly as the A-box deletion.
There was a correlation between the half-life of the
mutants and the degree of ubiquitination: while ubiqui-
tination of the ∆DAKTI mutant was reduced compared
for the more stable ∆A-box variant. Loss of ubiquitin-
ation seemed to affect Ste6 trafficking. In contrast
to wild-type Ste6, which was associated mainly with
mutants accumulated at the plasma membrane, as
demonstrated by immunofluorescence and cell frac-
tionation experiments. These findings suggest that
ubiquitination is required for efficient endocytosis of
Ste6 from the plasma membrane.
The yeast Ste6 protein is a typical member of the ABC-
transporter family consisting of 12 putative membrane-
(Kuchler et al., 1989; McGrath and Varshavsky, 1989).
By sequence comparison, it appears to be most closely
related to the mammalian Mdr proteins. This is under-
scored by the finding that mouse mdr3 is able to comple-
ment a ste6 defect in yeast (Raymond et al., 1992). Ste6
mutants are defective in the secretion of the mating
pheromone a-factor. Because of its similarity to other
proteins of the ABC-transporter family, which are known
to transport substrates across membranes, it has been
assumed that Ste6 acts as a transporter for a-factor at the
plasma membrane. We found, however, that most of
Ste6 is associated with internal membranes (Ko ¨lling and
Hollenberg, 1994). However, there is also evidence that
© Oxford University Press
R.Ko ¨lling and S.Losko
Fig. 2. Stability of Ste6 variants. Cells of strain JPY201, transformed
with different STE6 plasmids, were labeled with [35S]methionine for
15 min and were then chased with an excess of cold methionine. Ste6
was immunoprecipitated from cell extracts prepared at various time
intervals, as indicated. The precipitated proteins were analyzed by
SDS–PAGE and autoradiography. (A) JPY201/pRK257, (B) JPY201/
pRK264, (C) signal intensities of experiment A (s) and B (d) plotted
against the chase time, (D) JPY201/pRK265, (E) JPY201/pRK281,
(F) signal intensities of experiment D (u) and E (j) plotted against
the chase time.
Fig. 1. Predicted membrane topology of Ste6, structure of the linker
latory function. A role for ubiquitination other than target-
ing proteins for proteasomal degradation is further
proposed by Hicke and Riezman (1996), based on experi-
ments with the yeast α-factor receptor Ste2. Their data
indicate that ubiquitination triggers endocytosis of the
receptor–ligand complex leading to subsequent degrada-
tion in the vacuole. Other recent studies suggest that the
same may be true for other plasma membrane proteins
like uracil permease Fur4 (Volland et al., 1994), the
general amino acid permease Gap1 (Hein et al., 1995),
the multidrug permease Pdr5 (Egner et al., 1995; Egner
and Kuchler, 1996) and of course Ste6 (Berkower et al.,
1994; Ko ¨lling and Hollenberg, 1994).
A Ste6 mutant, which is no longer ubiquitinated, would
be an important tool to analyze the function of Ste6
the signal(s) in Ste6 required for ubiquitination. Here, we
show that ubiquitination and fast turnover of Ste6 are
mediated by a signal in the linker region connecting
the two ABC-transporter repeats. Our data indicate that
ubiquitination is important for efficient endocytosis of
tion box’. Upon closer inspection, we noticed an unusual
distribution of charged amino acids in the D-box region.
The ‘upstream’ half (amino acids 609–661) contains
mostly acidic amino acids (D,E ? 11; K ? 3) and was
therefore named ‘A-box’ for acidic box. The ‘downstream’
half (amino acids 662–716) contains two blocks of basic
amino acids (K,R ? 14; E ? 4) and was named ‘B-box’
for ‘basic box’. Furthermore, we found that the A-box
contains a sequence motif ‘DAKTI’, which closely
(Rohrer et al., 1993).
A-box, B-box and DAKTI deletion mutants were con-
structed in order to assess the influence of these regions
on the stability of Ste6. The half-lives of the Ste6 deletion
mutants were determined by pulse–chase experiments in
the STE6 deletion strain JPY201, transformed with Ste6-
encoding plasmids. Essentially the same results were
obtained with single-copy and multi-copy STE6 plasmids.
As can be seen from Figure 2, deletion of the A-box
strongly stabilized Ste6 [calculated half-life (τ) ? 68 min,
compared with 14 min for wild-type Ste6]. Deletion of
the B-box and the DAKTI sequence led to an ~2- to
3-fold stabilization of Ste6 (τ ? 43 and 36 min).
Recent data by Hicke and Riezman (1996) suggested
that the lysine residue in the DAKSS sequence of Ste2
acts as an acceptor for ubiquitin. Ubiquitination seems to
trigger endocytosis and subsequent degradation of Ste2 in
the vacuole. To test whether the lysine residue is crucial
for the destabilizing effect of the DAKTI sequence in Ste6,
we generated a lysine to arginine mutation (‘K612R’). This
exchange, however, had only a modest effect on the half-
life of Ste6 (τ ? 22 min, not shown). The K612R mutation
was constructed by site-directed mutagenesis of STE6,
Mutations in the linker region affect the half-life of
Ste6 is an extremely unstable protein with a half-life of
~14 min (Ko ¨lling and Hollenberg, 1994). Experiments
unrelated to the present study, where we generated chi-
meras between Ste6 and other ABC-transporters, indicated
that the 100 amino acid long linker region (amino acids
609–716) (Figure 1), which connects the two homologous
halves of Ste6, is important for the fast turnover. The
of strain JPY201, transformed with the different STE6
plasmids, by determining the frequency of zygote forma-
tion with a MATα tester strain. Again, the mating activity
was the same for wild-type Ste6 and the A-box and
DAKTI deletion variants and the K612R mutant (not
shown). No zygotes were observed for the vector control
and the B-box deletion. These results show that the A-box
and DAKTI deletion mutants are fully functional, in
The lack of activity and an aberrant intracellular localiz-
ation of the Ste6 ∆B-box protein (see below) suggest that
the stabilizing effect of the B-box deletion is indeed due
to misfolding of the protein and degradation of the protein
by a proteolytic pathway ‘slower’ than the normal Ste6
degradative pathway. The A-box and DAKTI deletions,
however, appear specifically to affect the turnover of Ste6.
Ste6 turnover is unaffected in proteasome
The finding that Ste6 is stabilized ~3-fold in a ubc4 ubc5
mutant (Seufert and Jentsch, 1990) suggested that the
ubiquitin system is involved in the degradation of Ste6
(Ko ¨lling and Hollenberg, 1994). Ubiquitination could
directly target Ste6 for degradation by the proteasome or
could have an indirect effect on the turnover of the
protein by affecting Ste6 trafficking. To test whether the
proteasome is involved directly in the degradation of Ste6,
we determined the half-life of Ste6 in proteasome mutants
by pulse–chase experiments. A congenic set of strains,
consisting of the wild-type strain WCG4a, the pre1-1
strain WCG4-11a and the pre1-1 pre2-2 strain WCG4-11/
22a (Heinemeyer et al., 1991, 1993), was transformed
with a multi-copy STE6 plasmid. Ste6 had about the same
half-life in all three strains (wild-type, 31 min; pre1-1, 27
min; pre1-1 pre2-2, 27 min; not shown). These data argue
against a role for the proteasome in Ste6 degradation. The
dramatic stabilization of Ste6 in a vacuolar pep4 mutant
(Berkower et al, 1994; Ko ¨lling and Hollenberg, 1994) also
suggests that the proteasome is not directly involved in
the degradation of Ste6.
Fig. 3. Halo-assay for a-factor activity. Serial dilutions (1:1, 1:2, 1:4
and 1:8) of culture supernatants of JPY201, transformed with different
single copy STE6 plasmids, were spotted onto a lawn of the a-factor
supersensitive sst2 strain XMW1-10D: (1) YEp420, (2) pRK109,
(3) pRK182, (4) pRK256, (5) pRK266, (6) pRK282, (7) pRK308,
(8) pRK309. a-factor activity is seen as a zone of growth inhibition.
while the other deletion mutants were constructed by the
insertion of PCR-generated cassettes into a modified STE6
gene that contained additional restriction sites flanking
the D-box-encoding sequences. The altered Ste6 protein
(Ste6*), encoded by this modified STE6 gene, behaved
like wild-type Ste6, with respect to function, localization
and turnover. To rule out that the additional restriction
sites introduced into STE6 contributed to the stabilizing
effect of the DAKTI deletion, a new deletion was con-
structed by site-directed mutagenesis which removes only
the DAKTI sequence. This ‘DAKTI*’ deletion was indis-
tinguishable from the original DAKTI deletion (τ ? 32
min, not shown).
The Ste6 ∆A-box mutant is no longer
Wild-type Ste6 accumulates in a highly ubiquitinated form
upon block of endocytosis (Ko ¨lling and Hollenberg, 1994).
We were interested to see whether the stabilization
observed with the ∆A-box and ∆DAKTI mutants was
correlated with a change in the ubiquitination pattern. To
facilitate detection of Ste6 ubiquitination, we made use
of a hemagglutinin (HA)-tagged ubiquitin variant. The
end4 ∆ste6 strain RKY592 was transformed with the
multi-copy plasmid YEp112 (Hochstrasser et al., 1991),
encoding the HA-tagged ubiquitin under the control of the
CUP1 promoter, and with multi-copy plasmids encoding
different Ste6 variants marked with a c-myc epitope tag.
The strains were grown for several generations at 25°C
in the presence of copper to induce expression of HA-
ubiquitin. After the cells had been shifted to restrictive
temperature (37°C) for 30 min to block endocytosis, cell
extracts were prepared. Proteins immunoprecipitated from
blotting with anti-myc antibodies to detect Ste6 and with
anti-HA antibodies to detect HA-ubiquitin covalently
The A-box and DAKTI deletion mutants are fully
To rule out that the observed stabilization is due to
misfolding of the mutant proteins, we measured the
a-factor transport activity in the STE6 deletion strain
JPY201, transformed with single-copy plasmids encoding
the different Ste6 variants. The amount of a-factor in the
culture supernatants was determined by a halo-assay.
Serial dilutions of the supernatants were spotted onto a
lawn of an a-factor supersensitive sst2 strain (Figure 3).
This strain has problems in recovering from pheromone-
induced cell cycle arrest due to a defect in the adaptation
response (Dietzel and Kurjan, 1987). A growth inhibitory
effect of culture supernatants comparable with wild-type
Ste6 was observed with the A-box and DAKTI deletion
mutants and the K612R mutant, i.e the supernatants of
these mutants contained a-factor activity in amounts
comparable with wild-type Ste6. No growth inhibition
was seen with the vector control and the B-box deletion.
In a second assay, we measured the mating activity
R.Ko ¨lling and S.Losko
Fig. 4. Ubiquitination of Ste6 variants. The end4 ∆ste6 strain RKY592
was transformed with YEp112, expressing an HA-tagged ubiquitin
variant (Hochstrasser et al., 1991), and with a plasmid expressing a
c-myc-tagged Ste6 variant. The STE6 plasmids were: (1) no plasmid,
(2) pRK257, (3) pRK264, (4) pRK281, (5) pRK310 and (6) pRK311.
The proteins immunoprecipitated from cell extracts with anti-Ste6
antibodies were analyzed by Western blotting with the anti-myc
antibody 9E10 (A). The blot was ‘stripped’ and reprobed with the
anti-HA antibody 12CA5 (B).
attached to Ste6. As can be seen from Figure 4A, all Ste6
variants were detectable in the anti-Ste6 immunoprecipit-
ates. Ubiquitination of Ste6 is detected as a high molecular
weight ‘smear’ on the anti-HA Western blot. A ubiquitin
smear was detected for all variants examined, except for
the ∆A-box mutant (Figure 4B). Thus, in contrast to wild-
type Ste6, the ∆A-box mutant does not seem to be
ubiquitinated. Also, we consistently observed that ubiquit-
ination was reduced compared with wild-type in the two
DAKTI deletion variants. The K612R mutant behaved
more or less like wild-type. An exact quantification of the
results is difficult, since the expression level of HA-
ubiquitin was somewhat variable between different experi-
ments due to fluctuations in copy number of the 2µ
vectors. However, in all experiments performed so far, we
never observed ubiquitination of the ∆A-box mutant and
we consistently saw less ubiquitination of the ∆DAKTI
mutants compared with wild-type Ste6.
Fig. 5. Localization of Ste6 variants by immunofluorescence. The
c-myc-tagged Ste6 variants, expressed in strain JPY201, were detected
with anti-myc primary antibodies and FITC-conjugated anti-mouse
secondary antibodies. STE6 plasmids: (A) pRK257, (B) pRK264,
(C) pRK281, (D) pRK310, (E) pRK311, (F) YEp420. Left panels:
FITC staining, right panels: phase contrast image.
Intracellular distribution of ubiquitination-deficient
Ubiquitination seems to act as an endocytosis signal for
the α-factor receptor Ste2 (Hicke and Riezman, 1996). To
of Ste6, we compared the intracellular distribution of the
ubiquitination-deficient Ste6 mutants with the distribution
of wild-type Ste6. The intracellular localization of myc-
tagged Ste6 variants, expressed from multi-copy plasmids,
was examined by immunofluorescence microscopy. Ste6
was detected with anti-myc antibodies and fluorescein
isothiocyanate (FITC)-conjugated secondary antibodies.
We noticed that the observed Ste6 immunofluorescence
staining was sensitive to the fixation conditions. Fixation
according to the standard protocol (Pringle et al., 1989)
for 2 h at 30°C gave rise to the patchy, Golgi-like staining
pattern, as reported previously (Ko ¨lling and Hollenberg,
1994); longer fixation (4 h, 30°C), however, revealed a
vacuolar staining (Figure 5A). The Ste6 staining sur-
rounded the vacuoles, which can be identified as white
spots in the phase contrast image. In some cells, a few
dots close to the vacuole were observed. These dots were
somewhat more obvious in the K612R mutant (Figure 5E).
In the ∆A-box (Figure 5B) and ∆DAKTI mutants
(Figure 5C and D), we observed a pronounced cell surface
staining, which was more intense in the bud than in the
mother cell. The surface staining was brighter in the ∆A-
box mutant than in the ∆DAKTI mutants. These data
show that the amount of Ste6 in the plasma membrane is
higher in the ∆A-box and ∆DAKTI deletion mutants
compared with wild-type.
To acquire more quantitative information about the
cell fractionation experiments. Cell extracts of JPY201
∆ste6, transformed with single-copy plasmids encoding
the different Ste6 variants, were fractionated on sucrose
density gradients. A thorough characterization of these
types of gradients has been presented in a previous report
(Ko ¨lling and Hollenberg, 1994). The plasma membrane
marker Pma1 (Serrano et al., 1986), which is very well
separated from internal membranes on these gradients,
was found mainly at high sucrose densities (fractions 15–
18, Figure 6A). The main peak of Ste6 was found in the
middle of the gradient around fraction 11 (Figure 6C).
Only a small amount of wild-type Ste6 could be detected
in the plasma membrane fraction. Essentially the same
pattern was seen with the ∆A-box and ∆DAKTI variants,
i.e. the main peak in fraction 11 and some Ste6 in the
plasma membrane fraction. However, it is clear that the
Fig. 7. Fractionation of Ste6 variants on sucrose gradients.
Densitometric quantification of the Ste6 signals in Figure 6.
Fig. 8. Fractionation of Ste6 by differential centrifugation. Cell
extracts were prepared from ConA-coated spheroplasts and centrifuged
at 3000 g for 15 min to pellet the plasma membranes (P3). The
supernatant was spun again at 100 000 g for 1 h to pellet the internal
membranes (P100). S100 ? supernatant from the 100 000 g spin.
Equal portions of the fractions were assayed for the presence of Ste6
(lanes 1–3) and Pma1 (lanes 4–6) by Western blotting. Extracts were
prepared from (A) JPY201 ∆ste6 ? pRK109 (Ste6*) and (B) JPY201
∆ste6 ? pRK256 (∆A-box).
Fig. 6. Fractionation of Ste6 variants on sucrose gradients. Whole cell
extracts of strain JPY201, transformed with different STE6 plasmids,
were fractionated by density gradient centrifugation (20–50% sucrose,
w/w, lowest density in fraction 1). Aliquots of the gradient fractions
were analyzed by SDS–PAGE and Western blotting. The STE6
plasmids were: (A, B and C) pRK109, (D) pRK256, (E) pRK282,
(F) pRK266. (A) Anti-Pma1 blot, (B) anti-Vph1 blot, (C–F) anti-Ste6
box and ∆DAKTI variants in the immunofluorescence
experiments looks impressive, the sucrose gradients show
only a comparatively small accumulation of the proteins
in the plasma membrane fraction. To obtain additional
information about the intracellular distribution of Ste6,
we prepared plasma membranes by a ‘ConA precipitation’
experiment (Patton and Lester, 1991). Spheroplasts were
coated with the lectin concanavalin A (ConA) which binds
to mannoproteins at the cell surface. ConA-stabilized
plasma membrane sheets can be recovered in the pellet
fraction after a low speed spin. The low speed supernatant
was spun again at 100 000 g for 1 h to pellet the remaining
cellular membranes. As can be seen from Figure 8A and
B, most of the plasma membrane marker Pma1 (76–90%)
was found in the low speed pellet (P3) while only a small
amount of Ste6* (13%) could be recovered in this fraction
(Figure 8A). The amount of Ste6 ∆A-box in the P3 pellet,
on the other hand, was substantially higher (Figure 8B).
More than half of the protein (57%) was found in
Taken together, our localization experiments show that
fraction of Ste6 ∆A-box (Figure 6D) and Ste6 ∆DAKTI
(Figure 6E) in the plasma membrane is significantly higher
than with wild-type Ste6. (A densitometric quantification
of these results is presented in Figure 7.) The Ste6 ∆B-
box mutant fractionated very differently from the other
variants (peak in fraction 6, Figure 6F), supporting the
view that it represents a misfolded protein which is
probably retained in the ER. The Ste6 distribution overlaps
but does not exactly coincide with the distribution of the
vacuolar marker Vph1 (Manolson et al., 1992) (peak in
fraction 10, Figure 6B). This result is compatible with the
view that part of Ste6, as suggested by the immunofluores-
cence experiments, is located in the vacuolar membrane.
Our experiments show that the ubiquitination-deficient
Ste6 mutants accumulate at the plasma membrane. The
immunofluorescence and sucrose gradient fractionation,
however, give a different picture regarding the magnitude
of this effect. While the surface staining of the ∆A-
R.Ko ¨lling and S.Losko
there is a substantial accumulation of the ubiquitin-
deficient mutants at the plasma membrane, indicating that
ubiquitination is important for efficient endocytosis of
Ste6. However, the fractionation experiments also show
that not all of the ∆A-box protein is present at the cell
surface. A substantial fraction of the protein still appears
to be associated with internal membranes.
The ubiquitination-deficient Ste6 mutants
accumulate at the plasma membrane upon block
One reason why the ubiquitination-deficient mutants do
not accumulate in large amounts at the plasma membrane
could be that they are impaired in their transport to the
plasma membrane. To test whether the mutants are able
to reach the plasma membrane, the end4 ∆ste6 strain
RKY592 was transformed with single-copy plasmids
encoding the Ste6 variants. The cells were grown at 25°C
for several generations and were then shifted for 1 h to
the restrictive temperature (37°C) to block endocytosis
(Raths et al., 1993). Cell extracts were prepared and
fractionated on sucrose gradients. Both, Ste6 ∆A-box
(Figure 9B) and Ste6 ∆DAKTI (Figure 9D) showed a
(fractions 15–18). The typical ‘ubiquitin-smear’ was seen
for wild-type Ste6 (Figure 9A). This ‘smear’ was reduced
in the ∆DAKTI mutant and was absent in the ∆A-box
mutant, in agreement with the results of the ubiquitination
experiment. The ∆B-box mutant did not accumulate in
the plasma membrane (Figure 9C). These results show
that the ubiquitination-deficient mutants are not impaired
in their transport to the plasma membrane. This is a further
indication that the deletion of the A-box and the DAKTI
sequence does not disturb the overall structure or function
of the protein.
In addition to the normal Ste6 band ?140 kDa, a
degradation band of ~55 kDa was observed. This band
was virtually absent in the ∆A-box mutant and was not
present at allin the ∆B-box mutant. It increasedin intensity
upon addition of trypsin (not shown). Thus, no specific
proteolytic activity seems to be required to generate this
Ste6 fragment. There is, however, a correlation between
the ubiquitination status of Ste6 and the appearance
of this band. One interpretation of this finding is that
ubiquitination leads to a conformational change in Ste6
fragment must be derived from the C-terminus of Ste6,
since the antibodies used to detect it were raised against
a C-terminal Ste6 fragment. A C-terminal fragment with
a calculated mol. wt of 55 kDa could be obtained by
cleavage in the third intracellular loop of Ste6. A similar
C-terminal proteolytic fragment has been described for
human Mdr1 (Yoshimura et al., 1989).
Fig. 9. Distribution of Ste6 variants in the endocytosis mutant end4.
Cell extracts of the end4 ∆ste6 strain RKY592, transformed with
different STE6 plasmids, were fractionated on sucrose gradients
(20–50% sucrose, w/w, lowest density in fraction 1). The cells were
grown at 25°C and shifted to 37°C for 1 h prior to extract preparation.
Aliquots of the gradient fractions were analyzed by SDS–PAGE and
Western blotting with anti-Ste6 antibodies. The STE6 plasmids were:
(A) pRK109, (B) pRK256, (C) pRK266, (D) pRK282.
A-box, B-box and D-box sequences. Pma1 is an extremely
stable protein with a reported half-life of 11 h (Benito
et al., 1991). The Ste6 sequences were fused to the
C-terminus of an HA-tagged variant of Pma1. The stability
of the fusions was analyzed by a gal-depletion experiment
(Figure 10). Strain JD52, expressing the different fusions
under the control of the GAL1 promoter, was pre-grown
on galactose medium, where the GAL1 promoter is active.
Then the cells were transferred to glucose medium to turn
off transcription from the GAL1 promoter. Samples were
taken at various time intervals and analyzed by Western
blotting with anti-HA antibodies. During the time course
of the experiment, the amount of HA-Pma1 (Figure 10A)
stayed more or less constant, in contrast to the amount of
the HA-Pma1–D-box fusion protein which was rapidly
diminished (Figure 10B). The estimated half-life of the
HA-Pma1–D-box fusion is ~30 min. The half-life of Pma1
was not significantly affected by the A-box (Figure 10C)
or the B-box (Figure 10D). This experiment shows that
the D-box not only destabilizes Ste6 but also the foreign
Pma1 protein. The A-box and B-box sequences alone,
The D-box affects stability and localization of
The A-box and DAKTI deletions could affect the turnover
of Ste6 either by removing a distinct degradation or
transport signal or by changing the overall structure of
the protein. If the D-box contains a distinct signal, it
should in principle be transferable to another membrane
protein. To test this prediction, we generated fusions
between the plasma membrane protein Pma1 and the
Fig. 10. Stability of HA-Pma1 fusions. Strain JD52, transformed with
different PMA1 plasmids, was grown on galactose medium to induce
expression of the Pma1 fusion proteins from the GAL1 promoter. At
time t ? 0 the cells were transferred to glucose medium. Cell extracts
were prepared at the time intervals indicated and analyzed by SDS–
PAGE and Western blotting with anti-HA antibodies. PMA1 plasmids:
(A) pRK315, (B) pRK318, (C) pRK319, (D) pRK320.
Fig. 11. Fractionation of Pma1 fusion proteins on sucrose gradients.
Whole-cell extracts of strain JD52, transformed with different PMA1
plasmids, were fractionated by density-gradient centrifugation
(20–50% sucrose, w/w, lowest density in fraction 1). The gradient
fractions were analyzed by SDS–PAGE and Western blotting with
anti-HA antibodies. The PMA1 plasmids were: (A) pRK315,
(B) pRK318, (C) pRK319, (D) pRK320.
however, were not sufficient to cause a destabilization
Next, we were interested to see whether the localization
of Pma1 was affected by the Ste6 sequences. The
intracellular distribution of the fusion proteins was ana-
lyzed by cell fractionation. Cell extracts from fusion
protein-expressing strains, grown on galactose medium,
were fractionated on sucrose gradients. Again, the main
portion of HA-Pma1 was found in the densest fractions
of the gradient (fractions 15–18, Figure 11A). However,
a substantial amount of HA-Pma1 was also observed in
the middle of the gradient. This differs from what we
have seen with wild-type Pma1 in glucose-grown cells
(Figure 6A), where the internal pool of Pma1 represented
only a minor fraction compared with the plasma membrane
pool (fractions 15–18). The distribution pattern of Pma1
may vary with growth conditions (galactose versus glu-
cose) or may be affected by the insertion of the HA-tag.
The important point, however, is that the distribution
of HA-Pma1–D-box (Figure 11B) is different from the
distribution of HA-Pma1 (Figure 11A). Most of HA-
Pma1–D-box is found in the internal pool and only a
small amount is found in the plasma membrane fraction.
Again, the complete D-box was required to see the effect,
since neither A-box (Figure 11C) nor B-box (Figure 11D)
alone affected the distribution of Pma1. The mostly
intracellular localization of the Pma1–D-box suggests that
the fusion proteinis internalized efficiently by endocytosis.
To exclude other possibilities, such as intracellular reten-
tion due to misfolding of the protein, it is important to
show that the protein still travels to the plasma membrane.
Fractionation experiments with an end4 mutant showed
that the fusion protein accumulates at the plasma mem-
brane upon block of endocytosis (not shown). The D-box,
therefore, seems to specifically affect the half-life of Pma1
by stimulating endocytosis.
The diffuse ‘smear’ on top of the HA-Pma1–D-box
fusion protein bands suggests that the fusion protein is
ubiquitinated. To test this prediction, an experiment similar
totheonedescribedinFigure 4wasperformed(Figure 12).
Strain JD52 was transformed with two plasmids, one
plasmid expressing either normal ubiquitin or c-myc-
tagged ubiquitin and another plasmid expressing either
HA-Pma1 or HA-Pma1–D-box. Proteins immunoprecipi-
tated from cell extracts with anti-HA antibodies were
assayed by Western blotting for the presence of HA-Pma1
with anti-HA antibodies (Figure 12A) and for the presence
of c-myc-tagged ubiquitin, covalently attached to Pma1,
with anti-c-myc antibodies (Figure 12B). On the c-myc
blot a strong signal was detected for the HA-Pma1–D-
box protein (Figure 12B, lane 4), demonstrating that this
protein is ubiquitinated. A faint signal was also detected
for the HA-Pma1 protein (Figure 12B, lane 2). HA-Pma1
alone, therefore, seems to be already ubiquitinated to a
certain degree. This ubiquitination, however, is drastically
enhanced by the addition of the D-box. No signals were
obtained for the controls expressing only normal ubiquitin
(Figure 12B, lanes 1 and 3).
R.Ko ¨lling and S.Losko
Table I. STE6 plasmids
Plasmid Ste6 variant Mutant typeVectora
Fig. 12. Ubiquitination of the HA-Pma1–D-box fusion. Strain JD52
was transformed with a plasmid expressing either normal ubiquitin
(YEp96) or c-myc-tagged ubiquitin (YEp101–) (Hochstrasser et al.,
1991) and with a plasmid expressing either HA-tagged Pma1
(pRK315) or the HA-tagged Pma1–D-box fusion (pRK318). The
plasmid combinations were: (1) YEp96, pRK315; (2) YEp105,
pRK315; (3) YEp96, pRK318; (4) YEp105, pRK318. Proteins
immunoprecipitated from cell extracts with anti-HA antibodies were
analyzed by Western blotting with anti-HA antibodies (A). The blot
was ‘stripped’ and reprobed with anti-c-myc antibodies (B).
apRK182 and pRK278 ? single-copy plasmids, pYKS1 ? multi-copy
A-box alone was not sufficient to destabilize Pma1. The
importance of the B-box for Ste6 turnover could not be
evaluated by deletion analysis, since deletion of the B-box
gave rise to a non-functional and probably misfolded
protein. This is not totally unexpected, since deletion of the
B-box places the A-box, which contains many negatively
charged amino acids, just upstream of the first transmem-
brane span of the second half of Ste6. It has been shown
that a net positive charge just upstream of a membrane-
spanning segment is required for a Nin–Coutorientation
(Gafvelin and von Heijne, 1994, a net negative charge
in this position may, therefore, inverse the membrane
orientation of the membrane span, which would change
the membrane topology of the protein.
The Ste6 turnover signal
In this study, we show that the Ste6 linker region contains
a signal which mediates ubiquitination and fast turnover
of Ste6. Which part of the linker region forms the signal?
The linker region is composed of an acidic half (A-box)
and a basic half (B-box). Deletion of the acidic half has
a strong impact on the stability of Ste6 and completely
prevents ubiquitination of the protein. The A-box, there-
fore, either contains the turnover signal or is part of a
larger signal which is inactivated upon deletion of the
A-box. A potential candidate for a turnover signal is the
sequence motif ‘DAKTI’ found in the A-box region. This
sequence resembles the ‘DAKSS’ signal shown to be
required for the endocytosis of a truncated Ste2 receptor
(Rohrer et al., 1993). Deletion of the DAKTI sequence
indeed affects the turnover of Ste6. The stabilizing effect,
however, is not as strong as the effect of the A-box
deletion, indicating that the DAKTI sequence is only part
of the signal. Alternatively, additional redundant signals
may exist in the A-box region.
The lysine residue in the Ste2 DAKSS sequence has
been shown to be essential for the internalization of a
C-terminally truncated receptor (Rohrer et al., 1993; Hicke
and Riezman, 1996). An exchange of this lysine residue
for arginine eliminates ubiquitination of the truncated
receptor, supporting the view that ubiquitination triggers
endocytosis of Ste2. A lysine to arginine mutation in the
Ste6 DAKTI sequence, however, had only a minor effect
on the Ste6 turnover (τ ? 14 min → τ ? 22 min).
Also, ubiquitination of Ste6 did not seem to be affected.
However, we still cannot exclude that this lysine functions
as an acceptor for ubiquitin. Apparently, the ubiquitination
machinery seems to be able to use a number of alternative
lysine residues as acceptors in the vicinity of a degradation
signal (Kornitzer et al., 1994). This seems to be true even
for the Ste2 receptor, since the effect of the lysine mutation
in the DAKSS sequence has only been demonstrated for
a truncated version of Ste2 where the C-terminus, which
contains several lysines, has been deleted.
That the B-box also contributes to ubiquitination and
turnover of Ste6 is suggested by the analysis of the Pma1–
Ste6 fusions. In contrast to the D-box, which caused a
marked destabilization and ubiquitination of Pma1, the
The role of Ste6 ubiquitination
Analysis of the different Ste6 variants established a
correlation between the Ste6 half-life and the extent of
ubiquitination. This correlation could be explained by
postulating a role for ubiquitination in targeting Ste6 for
degradation bythe proteasome. Theproteasome has indeed
been implicated in the degradation of another ABC-
transporter, the CFTR protein (Jensen et al., 1995; Ward
et al., 1995). However, the finding that the half-life of
Ste6 was unaffected by mutations in the proteasome
subunits Pre1 and Pre2 (Heinemeyer et al., 1991, 1993)
argues against a role for the proteasome in Ste6 degrada-
tion. Moreover, previous experiments, which point to the
vacuole as the major site of Ste6 degradation (Berkower
et al., 1994; Ko ¨lling and Hollenberg, 1994), provide a
strong argument against a direct involvement of the
proteasome in Ste6 turnover.
As described for Ste2, ubiquitination may also indirectly
affect the turnover of Ste6 by inducing endocytosis and
subsequent degradation in the vacuole (Hicke and Riez-
man, 1996). Our localization experiments show that the
ubiquitination-deficient Ste6 mutants accumulate at the
plasma membrane, suggesting that ubiquitination indeed
stimulates endocytosis of Ste6. The different methods
used to assess the cellular distribution of the Ste6 mutants
give somewhat different results regarding the magnitude
of this effect. The immunofluorescence experiment and
has a strong impact on the internalization of Ste6. The
sucrose gradients, on the other hand, show only a compar-
Table II. Yeast strains
Strain Genotype Source
MATα ura3-52 his4-619
MATa ura3-52 his3-∆200 leu2-3,112 trp1-∆63 lys2-801
MATa ste6-∆1::HIS3 gal2 his3-∆200 leu2-3,112 lys2-801 trp1-1 ura3-52
ura3 trp1ste6-∆1::HIS3 end4 his3 leu2 gal
J.Dohmen, Du ¨sseldorf
J.McGrath (McGrath and Varshavsky, 1989)
backcrossed from strain RH 268-1C
(Raths et al., 1993)
MATa ura3 leu2-3,112 his3-11,15 CANs
MATa pre1-1 ura3 leu2-3,112 his3-11,15 CANs
MATa pre1-1 pre2-2 ura3 leu2-3,112 his3-11,15 CANs
MATα cry1-11 thr4 ssl2-1
Site-directed mutagenesis was performed with the Bio-Rad Muta-
Gene™ kit based on the method of Kunkel et al. (1987). A 1.2 kb
internal PstI STE6 fragment (position 1620–2830) was mutagenized with
the following mutagenic primers (exchanges marked in bold print):
∆DAKTI*, 5?-C (position 2278) TACAGAATGACTACTCTGTCG-
ACACAGAGACTGAAGAAAAATC; K612R, 5?-G (position 2283)
AAG. The mutations were confirmed by sequencing.
The 1.2 kb internal PstI fragments, modified by mutagenesis or by
insertion of the PCR cassettes, were cloned into pRK182, pRK278 and
pYKS1 (Kuchler et al., 1993), as summarized in Table I. The plasmids
pRK182 and pRK278 are based on the single-copy vectors YCp50 (Rose
et al., 1987) and YCplac33 (Gietz and Sugino, 1988) and contain a
6.2 kb BglII–SalI STE6 fragment. pYKS1 is a multi-copy vector, based
on YEp352, containing a c-myc-tagged STE6 gene (Kuchler et al., 1993).
Plasmid pRK315, based on the single copy vector YCp50, encodes a
modified PMA1 gene under the control of the GAL1 promoter. The N-
terminal part of PMA1, containing an HA tag, was derived from pXZ28
(kindly provided by Jim Haber). Two modifications were introduced into
the C-terminal part, which was generated by assembly PCR. First, the
internal BamHI site (position 2535; Serrano et al., 1986) was removed,
without changing the amino acid sequence and, second, a new BamHI
site was placed behind the last codon [5?-...GAAACC (position 3690)
GGATCC-3?]. Furthermore, stop codons in all three reading frames were
placed downstream from the BamHI site. In-frame fusions with the STE6
D-box (pRK318), A-box (pRK319) and B-box (pRK320) were generated
by inserting the PCR fragments described above into the C-terminal
BamHI site of the modified PMA1 gene.
The yeast strains used are listed in Table II.
atively small accumulation of the ubiquitination-deficient
proteins in the plasma membrane fraction. The plasma
membrane fraction on the sucrose gradients, however, is
only operationally defined as the fraction where Pma1,
the plasma membrane ATPase, is located. It is conceivable
that the plasma membrane consists of subdomains which
do not fractionate together. It could well be possible,
therefore, that a fraction of Ste6 is localized to plasma
membrane subdomains, e.g. domains actively involved in
endocytosis, which do not contain Pma1. Separation of
plasma membrane subdomains should not occur in the
ConA precipitation experiment. Therefore, the ConA pre-
cipitation experiment would be expected to give a more
accurate picture of the amount of Ste6 at the cell surface.
Although all localization experiments show that the
ubiquitination-deficient Ste6 ∆A-box mutant accumulates
at the cell surface, it is also clear that a substantial fraction
of the protein is still associated with internal membranes.
If endocytosis of the mutant protein from the plasma
membrane were the only rate-limiting step in the degrada-
tion pathway, most of Ste6 ∆A-box should be found
in the plasma membrane. There are several possible
explanations why not all of the ubiquitination-deficient
Ste6 protein is found in the plasma membrane. One
explanation is that endocytosis still occurs in the absence
is that not all of the Ste6 protein reaches the vacuole via
the plasma membrane. A certain fraction of the protein
could travel directly from the Golgi to the vacuole via a
pre-vacuolar compartment. The finding that Ste6 is stabil-
ized only 3-fold in late secretory mutants is in line
with this interpretation. A third possibility is that Ste6
ubiquitination stimulates another internal transport step
which becomes rate-limiting for the Ste6 ∆A-box protein
due to the absence of ubiquitination.
Pulse–chase experiments and immunoprecipitation
The procedure for the pulse–chase experiments and immunoprecipitation
has been described in detail previously (Ko ¨lling and Hollenberg, 1994).
Aliquots corresponding to µ 1 OD600unit cells were removed at 20 min
intervals and lysed by vortexing with glass beads for 3 min. The relative
amount of Ste6 at each time point was quantified by scanning of
autoradiograms with a Howtek Scanmaster 3 scanner and analysis with
the Imagemaster 1D software.
Quantitative mating assay and halo-assay
To determine the mating activity, equal numbers (1?107) of the exponen-
tially growing MATa strain JPY201, transformed with different Ste6
plasmids, and the MATα tester strain DBY2058 were mixed. The cells
were pelleted, overlayed with 5 ml of YPD medium and incubated for
4 h at 30°C. Aliquots of serial dilutions were plated on rich medium, to
determine the total cell number and on selective plates allowing only
for growth of the zygotes. The fraction of cells which were able to grow
for a wild-type MATa mating partner was usually ~10%. For the ∆ste6
MATa strain JPY201 no zygotes were detected among 5?106cells
For the halo-assay, overnight cultures of JPY201, transformed with
different Ste6 plasmids, were diluted to an OD600? 0.2 in SD ?
casamino acids medium and were grown for 4 h at 30°C. The culture
supernatants were spotted onto a lawn of sst2 (? ssl2) cells (2?106
cells of strain XMW1-10D in 3 ml of YPD top agar poured onto a YPD
plate) using a 32-point inoculator (‘frogger’).
Materials and methods
Plasmids and yeast strains
To facilitate the construction of STE6 linker mutants, additional BamHI
sites were introduced by PCR at the end of repeat 1 of STE6, 5?-
...AATGAC (position 2289) GGATCC-3? and at the beginning of repeat
2 of STE6, 5?-GGATCC (position 2617) TAATC...-3? (for sequence
positions see Kuchler et al., 1989; the original STE6 sequences are in
bold print). The following PCR-generated cassettes with one BamHI and
one BglII end were inserted between the two BamHI sites. Ste6*, 5?-
GGATCCATAATGT (position 2293) CTGAT... CAAAAA (position
2612) GATCT-3?; ∆A-box, 5?-GGATCCC (position 2452) AAAAA...
CAAAAA (position 2612) GATCT-3?; ∆B-box, 5?-GGATCCATAATGT
(position 2293) CTGAT... GCAATC (position 2448) CAGATCT-3?;
∆DAKTI, 5?-GGATCCG (position 2311) TAGAT... CAAAAA (position
The immunofluorescence experiments were performed essentially as
described in Pringle et al. (1989). Cells were grown to exponential
R.Ko ¨lling and S.Losko
phase (A600? 0.5–0.8, 3–4?107/ml) and fixed directly for 4 h with
formaldehyde (final concentration 5%). The fixed cells were sphero-
plasted and extracted with 0.1% Triton X-100 for 5 min and then attached
to a multiwell slide treated with 0.1% polylysine (Sigma). The cells
were first incubated with the anti-c-myc mouse monoclonal primary
antibody (9E10, Berkeley Antibody Co, Inc.) (1:200 dilution in PBS ?
1 mg/ml BSA) for 90 min and then another 90 min with FITC-conjugated
anti-mouse secondary antibodies (Dianova, 1:300 dilution in PBS/BSA).
Finally, the cells were incubated for 5 min with 4,6 -diamidino-2-
phenylindole (DAPI) (1 mg/ml). The cells were examined with a Zeiss
Axioskop and photographed with Ilford FP4 black and white film.
Multiple ubiquitin-conjugating enzymes participate in the in vivo
degradation of the yeast MATα2 repressor. Cell, 74, 357–369.
Ciechanover,A. (1994) The ubiquitin–proteasome proteolytic pathway.
Cell, 79, 13–21.
Deshaies,R.J., Chau,V. and Kirschner,M. (1995) Ubiquitination of the
G1 cyclin Cln2p by a Cdc34p-dependent pathway. EMBO J., 14,
Dietzel,C. and Kurjan,J. (1987) Pheromonal regulation and sequence of
the Saccharomyces cerevisiae SST2 gene: a model for desensitization
to pheromone. Mol. Cell. Biol., 7, 4169–4177.
Egner,R. and Kuchler,K. (1996) The yeast multidrug transporter Pdr5 of
the plasma membrane is ubiquitinated prior to endocytosis and
degradation in the vacuole. FEBS Lett., 378, 177–181.
Egner,R., Mahe,Y., Pandjaitan,R. and Kuchler,K. (1995) Endocytosis
and vacuolar degradation of the plasma membrane-localized Pdr5
cerevisiae. Mol. Cell. Biol., 15, 5879–5887.
Gafvelin,G. and von Heijne,G. (1994) Topological ‘frustration’ in
multispanning E.coli inner membrane proteins. Cell, 77, 401–412.
Gietz,R.D. and Sugino,A. (1988) New yeast–Escherichia coli shuttle
vectors constructed with in vitro mutagenized yeast genes lacking six-
base pair restriction sites. Gene, 74, 527–534.
Hein,C., Springael,J.Y., Volland,C., Haguenauer,T.R. and Andre,B.
(1995) NPl1, an essential yeast gene involved in induced degradation
of Gap1 and Fur4 permeases, encodes the Rsp5 ubiquitin–protein
ligase. Mol. Microbiol., 18, 77–87.
multicatalytic–multifunctional proteinase: mutants unravel its function
in stress induced proteolysis and uncover its necessity for cell survival.
EMBO J., 10, 555–562.
Heinemeyer,W., Gruhler,A., Mohrle,V., Mahe,Y. and Wolf,D.H. (1993)
PRE2, highly homologous to the human major histocompatibility
complex-linked RING10 gene, codes for a yeast proteasome subunit
necessary for chrymotryptic activity and degradation of ubiquitinated
proteins. J. Biol. Chem., 268, 5115–5120.
Hicke,L. and Riezman,H. (1996) Ubiquitination of a yeast plasma
membrane receptor signals its ligand-stimulated endocytosis. Cell, 84,
Hochstrasser,M. (1995) Ubiquitin, proteasomes, and the regulation of
intracellular protein degradation. Curr. Opin. Cell Biol., 7, 215–223.
Hochstrasser,M., Ellison,M.J., Chau,V. and Varshavsky,A. (1991) The
short-lived MATα2 transcriptional regulator is ubiquitinated in vivo.
Proc. Natl Acad. Sci. USA, 88, 4606–4610.
Jensen,T.J., Loo,M.A., Pind,S., Williams,D.B., Goldberg,A.L. and
Riordan,J.R. (1995) Multiple proteolytic systems, including the
proteasome, contribute to CFTR processing. Cell, 83, 129–135.
Jentsch,S. and Schlenker,S. (1995) Selective protein degradation: a
journey’s end within the proteasome. Cell, 82, 881–884.
Kaplon,T. and Jacquet,M. (1995) The cellular content of Cdc25p, the
Ras exchange factor in Saccharomyces cerevisiae, is regulated by
destabilization through a cyclin destruction box. J. Biol. Chem., 270,
Ko ¨lling,R. and Hollenberg,C.P. (1994) The ABC-transporter Ste6
accumulates in the plasma membrane in a ubiquitinated form in
endocytosis mutants. EMBO J., 13, 3261–3271.
Kornitzer,D., Raboy,B., Kulka,R.G. and Fink,G.R. (1994) Regulated
degradation of the transcription factor Gcn4. EMBO J., 13, 6021–6030.
Kuchler,K., Sterne,R.E. and Thorner,J. (1989) Saccharomyces cerevisiae
STE6 gene product: a novel pathway for protein export in eukaryotic
cells. EMBO J., 8, 3973–3984.
(STE6 gene product) and cell polarity in the yeast Saccharomyces
cerevisiae. J. Cell Biol., 120, 1203–1215.
Kunkel,T.A., Roberts,J.D. and Zakour,R.A. (1987) Rapid and efficient
site-specific mutagenesis without phenotypic selection. Methods
Enzymol., 154, 367–382.
Madura,K. and Varshavsky,A. (1994) Degradation of G alpha by the
N-end rule pathway. Science, 265, 1454–1458.
Manolson,M.F. et al. (1992) The VPH1 gene encodes a 95-kDa integral
membrane polypeptide required for in vivo assembly and activity of
the yeast vacuolar H(?)-ATPase. J. Biol. Chem., 267, 14294–14303.
McGrath,J.P. and Varshavsky,A. (1989) The yeast STE6 gene encodes a
homologue of the mammalian multidrug resistance P-glycoprotein.
Nature, 340, 400–404.
The sucrose density gradient fractionation was performed, according to
a protocol obtained from R.Serrano, as described previously (Ko ¨lling
and Hollenberg, 1994).
The ConA precipitation experiment is based on a protocol by Patton
and Lester (1991). Forty A600units of cells from an exponentially
growing culture (A600? 0.4, 2?107/ml) were harvested, washed in
H2O and incubated in pre-treatment buffer (100 mM EDTA, 0.5%
β-mercaptoethanol, 10 mM Tris–HCl, pH 7.5) at 4 ml/g cells (wet
weight) for 20 min at 30°C. Then 10 mM NaN3was added to the
culture and subsequently to the buffers to stop ATP-dependent transport
processes. The cells were then washed and resuspended in S buffer
(1.2 M sorbitol, 0.5 mM MgCl2, 40 mM HEPES, pH 7.5) to 4 ml/g cells
and spheroplasted by the addition of 50 µg of oxalyticase (Enzogenetics)
for 45–60 min at 30°C. Spheroplasts were centrifuged through a 5 ml
cushion of 1.7? S buffer for 5 min at 500 g. The spheroplasts were
washed in 5 ml of S buffer and in 5 ml of ConA buffer (1.2 M sorbitol,
1 mM Mg acetate, 1 mM CaCl2, 1 mM MnCl2, 50 mM Tris–HCl, pH 7.5)
and were carefully resuspended in ConA buffer at 20 ml/g cells. One
volume of ConA buffer containing ConA (3.5 mg/g cells) was added
dropwise while gently stirring at room temperature. After addition of
ConA, the stirring was continued for another 10 min. Then the sphero-
plasts were harvested by a 5 min centrifugation at 500 g, washed twice
in ConA buffer and resuspended in cold lysis buffer [2 mM EDTA,
1 mM dithiothreitol (DTT), 5 mM HEPES, pH 7.5] at 50 ml/g cells. The
lysis buffer contained the following protease inhibitors: 0.5 µg/ml each
of aprotinin, antipain, chymostatin, leupeptin, pepstatin A, 1.6 µg/ml
benzamidine, 1 µg/ml phenanthroline, 170 µg/ml phenylmethylsulfonyl
fluoride. The following steps were performed at 4°C. The spheroplasts
were lysed with 20 strokes in a Dounce homogenizer using a tight fitted
pestle. One ml of the homogenate was transferred to a microcentrifuge
tube and centrifuged twice for 5 min at 100 g. The supernatant was
centrifuged for 15 min at 3000 g. The plasma membrane-containing
pellet (P3) was then resuspended in 500 µl of storage buffer (20%
glycerol, 1 mM DTT, 0.1 mM EDTA, 10 mM Tris–HCl, pH 7.5). The
supernatant was again centrifuged for 1 h at 100 000 g (45 000 r.p.m.
in a Beckman TLA45 rotor). The pellet (P100) was again resuspended
in storage buffer.
Plasmids and strains were generously provided by Ju ¨rgen Dohmen, Jim
Haber, Wolfgang Heinemeyer, Mark Hochstrasser, Karl Kuchler, Linda
Hicke, Howard Riezman and Wolfgang Zachariae. We thank Frauke
Bu ¨hring and Andreas Kranz for their help in plasmid construction. We
are also grateful to Ju ¨rgen Dohmen and Karl Kuchler for stimulating
discussions and to Cor Hollenberg for his support. This work was
by the ‘Biologisch-Medzinisches Forschungszentrum (BMFZ)’ at the
University of Du ¨sseldorf and by the DFG grant Ko 963/2-1 to R.K.
Benito,B., Moreno,E. and Lagunas,R. (1991) Half-life of the plasma
membrane ATPase and its activating system in resting yeast cells.
Biochim. Biophys. Acta, 1063, 265–268.
Berkower,C., Loayza,D. and Michaelis,S. (1994) Metabolic instability
and constitutive endocytosis of STE6, the a-factor transporter of
Saccharomyces cerevisiae. Mol. Biol. Cell, 5, 1185–1198.
Biederer,T., Volkwein,C. and Sommer,T. (1996) Degradation of subunits
of the Sec61p complex, an integral component of the ER membrane,
by the ubiquitin–proteasome pathway. EMBO J., 15, 2069–2076.
Chen,P., Johnson,P., Sommer,T., Jentsch,S. and Hochstrasser,M. (1993)
Ste6 ubiquitination Download full-text
multiubiquitination and deubiquitination of high-affinity immuno-
globulin E receptors. EMBO J., 12, 779–786.
Patton,J.L. and Lester,R.L. (1991) The phosphoinositol sphingolipids of
Saccharomyces cerevisiae are highly localized in the plasma
membrane. J. Bacteriol., 173, 3101–3018.
Pringle,J.R., Preston,R.A., Adams,A.E.M., Stearns,T., Drubin,D.G.,
Haarer,B.K. and Jones,E.W. (1989) Fluorescence microscopy methods
for yeast. Methods Cell Biol., 31, 357–434.
Raths,S., Rohrer,J., Crausaz,F. and Riezman,H. (1993) end3 and end4:
in Saccharomyces cerevisiae. J. Cell Biol., 120, 55–65.
Raymond,M., Gros,P., Whiteway,M. and Thomas,D.Y. (1992) Functional
complementation of yeast ste6 by a mammalian multidrug resistance
mdr gene. Science, 256, 232–233.
Rohrer,J., Be ´ne ´detti,H., Zanolari,B. and Riezman,H. (1993) Identification
of a novel sequence mediating regulated endocytosis of the G protein-
coupled alpha-pheromone receptor in yeast. Mol. Biol. Cell, 4,
Rose,M.D., Novick,P., Thomas,J.H., Botstein,D. and Fink,G.R. (1987)
A Saccharomyces cerevisiae genomic plasmid bank based on a
centromere-containing shuttle vector. Gene, 60, 237–243.
Schork,S.M., Thumm,M. and Wolf,D.H. (1995) Catabolite inactivation of
fructose-1,6-bisphosphatase of Saccharomyces cerevisiae. Degradation
occurs via the ubiquitin pathway. J. Biol. Chem., 270, 26446–26450.
Serrano,R., Kielland-Brandt,M.C. and Fink,G.R. (1986) Yeast plasma
membrane ATPase is essential for growth and has homology with
(Na?? K?), K?- and Ca2?-ATPases. Nature, 319, 689–693.
Seufert,W. and Jentsch,S. (1990) Ubiquitin-conjugating enzymes UBC4
and UBC5 mediate selective degradation of short-lived and abnormal
proteins. EMBO J., 9, 543–550.
Seufert,W., Futcher,B. and Jentsch,S. (1995) Role of a ubiquitin-
conjugating enzyme in degradation of S- and M-phase cyclins. Nature,
Endocytosis and degradation of the yeast uracil permease under
adverse conditions. J. Biol. Chem., 269, 9833–9841.
Ward,C.L., Omura,S. and Kopito,R.R. (1995) Degradation of CFTR by
the ubiquitin–proteasome pathway. Cell, 83, 121–127.
degradation. Mol. Cell. Biol., 15, 731–741.
Yoshimura,A., Kuwazuru,Y., Sumizawa,T., Ichikawa,M., Ikeda,S.-i.,
Uda,T. and Akiyama,S.-i. (1989) Cytoplasmic orientation and two-
domain structure of the multidrug transporter, P-glycoprotein,
demonstrated with sequence-specific antibodies. J. Biol. Chem., 264,
andKinet,J.P. (1993)Cellsurfacecontrol ofthe
Received on July 8, 1996; revised on December 31, 1996