Tải bản đầy đủ (.pdf) (14 trang)

Tài liệu Báo cáo khoa học: A genetic screen identifies mutations in the yeastWAR1 gene, linking transcription factor phosphorylation to weak-acid stress adaptation docx

Bạn đang xem bản rút gọn của tài liệu. Xem và tải ngay bản đầy đủ của tài liệu tại đây (833.05 KB, 14 trang )

A genetic screen identifies mutations in the yeast WAR1
gene, linking transcription factor phosphorylation
to weak-acid stress adaptation
Christa Gregori*, Bettina Bauer*, Chantal Schwartz, Angelika Krenà, Christoph Schu
¨
ller§
and Karl Kuchler
Medical University Vienna, Max F. Perutz Laboratories, Department of Medical Biochemistry, Campus Vienna Biocenter, Austria
Weak acids have a long history as additives in food
preservation. In addition to sulfites used in wine mak-
ing, acetic, sorbic, benzoic and propionic acids are
commonly used in the food and beverage industry
to prevent spoilage [1,2]. In solution, weak acids exist
in a dynamic equilibrium between undissociated,
uncharged molecules and their anionic form. These
acids display increased antimicrobial action at low pH,
which favors the undissociated state. The uncharged
molecules can readily diffuse through the plasma
Keywords
ABC transporter; stress response; weak
organic acids; yeast; zinc finger
Correspondence
K. Kuchler, Medical University Vienna, Max
F. Perutz Laboratories, Department of
Medical Biochemistry, Campus Vienna
Biocenter, Dr Bohr-Gasse 9 ⁄ 2, A-1030,
Vienna, Austria
Fax: +43 1 4277 9618
Tel: +43 1 4277 61807
E-mail:
Present address


Universite
´
de Nice-Sophia Antipolis,
Inserm, U636, Centre de Biochimie, UFR
Sciences, Parc Valrose, Nice, France
àInstitute of Biochemistry and Genetics,
Department of Clinical and Biological
Research (DKBW), Basel, Switzerland
§University of Vienna, Max F. Perutz
Laboratories, Department of Biochemistry &
Molecular and Cellular Biology, Campus
Vienna Biocenter, Austria
*
These authors contributed equally to this
work
(Received 11 January 2007, revised 4 April
2007, accepted 19 April 2007)
doi:10.1111/j.1742-4658.2007.05837.x
Exposure of the yeast Saccharomyces cerevisiae to weak organic acids such
as the food preservatives sorbate, benzoate and propionate leads to the
pronounced induction of the plasma membrane ATP-binding cassette
(ABC) transporter, Pdr12p. This protein mediates efflux of weak acid ani-
ons, which is essential for stress adaptation. Recently, we identified War1p
as the dedicated transcriptional regulator required for PDR12 stress
induction. Here, we report the results from a genetic screen that led to the
isolation of two war1 alleles encoding mutant variants, War1-28p and
War1-42p, which are unable to support cell growth in the presence of sorb-
ate. DNA sequencing revealed that War1-28 encodes a truncated form of
the transcriptional regulator, and War1-42 carries three clustered mutations
near the C-terminal activation domain. Although War1-42 is expressed and

properly localized in the nucleus, the War1-42p variant fails to bind the
weak-acid-response elements in the PDR12 promoter, as shown by in vivo
footprinting. Importantly, in contrast with wild-type War1p, War1-42pis
also no longer phosphorylated upon weak-acid challenge, demonstrating
that phosphorylation of War1p, its activation and DNA binding are tightly
linked processes that are essential for adaptation to weak-acid stress.
Abbreviations
GST, glutathione S-transferase; MHR, middle homology region; NLS, nuclear localization signal; PDR, pleiotropic drug resistance; WARE,
weak-acid-response element; YPD, yeast peptone ⁄ dextrose.
3094 FEBS Journal 274 (2007) 3094–3107 ª 2007 The Authors Journal compilation ª 2007 FEBS
membrane. In the cytoplasm, weak acids encounter a
more neutral pH, causing their dissociation into acid
anions and protons. The protons lead to cytoplasmic
acidification, thereby inhibiting important metabolic
processes such as glycolysis [3], possibly interfering
with active transport and signal transduction [1]. Fur-
thermore, sorbate and benzoate may also act as
membrane-damaging substances [4] and, at least under
aerobic conditions, cause severe oxidative stress [5,6].
The antimicrobial action of weak-acid preservatives
is usually characterized by extended lag phases and cell
stasis, although microbial killing does not occur. How-
ever, cells can adapt to the presence of weak acid and
resume growth. In Saccharomyces cerevisiae, this adap-
tation requires induction of the Pdr12p plasma mem-
brane ATP-binding cassette (ABC) transporter [7].
Together with the plasma membrane H
+
-ATPase,
Pma1p, the activity of which is also regulated by weak

acid stress [8,9], Pdr12p becomes one of the most abun-
dant surface proteins in stressed cells [7]. Whereas
Pma1p effluxes protons, Pdr12p mediates cellular
extrusion of weak acid anions [7]. Notably, other mem-
bers of the fungal ABC transporter family transport a
wide variety of different xenobiotics across the plasma
membrane or membranes of subcellular compartments
[10,11]. Pdr12p is the essential component of this stress
response pathway, as cells are hypersensitive to sorbic,
benzoic and propionic acid [7] and fail to adapt to such
stress conditions in the absence of Pdr12p. Moreover,
recent data indicate the involvement of Pdr12p in the
export of by-products of amino-acid catabolism, as a
pdr12D strain displays hypersensitivity to fusel acids
derived from leucine, isoleucine, valine, phenylalanine
and tryptophan [12]. Therefore, Pdr12p is not only
required for adaptation to weak-acid stress, but might
also efflux weak-acid metabolites. Notably, PDR12 is
rapidly induced by weak-acid challenge [7], but also in
cultures grown with leucine, methionine or phenylalan-
ine as sole nitrogen source [12]. A recent study [13]
attempted to identify Pdr12p-like proteins in other food
spoilage yeasts. A sorbic-inducible protein cross-react-
ing with S. cerevisiae Pdr12p antibodies in Saccharomy-
ces bayanus was found. In contrast, proteins detectable
with the same antibodies in Zygosaccharomyces bailii
and Zygosaccharomyces lentus were not up-regulated
upon sorbate challenge.
We are interested in identifying components of the
signaling pathway required for this efficient response in

S. cerevisiae. Hence, we pursued two different strategies.
First, we used a functional genomics approach and
screened all putative nonessential transcription factor
deletions of the EUROSCARF collection [14] (http://
www.uni-frankfurt.de/fb15/mikro/eroscarf/) for sorbate
hypersensitivity. This approach identified the regulator
War1p (weak acid resistance) as the main inducer of
PDR12 [15]. War1p is a nuclear transcription factor,
which decorates at least one weak-acid-response element
(WARE) in the PDR12 promoter. War1p is rapidly
phosphorylated upon stress challenge, and phosphory-
lation is somehow coupled to War1p activation [15].
Interestingly, War1p is required for PDR12 up-regu-
lation in response to exogenous weak-acid stress, but it
appears also to be involved in the metabolism-derived
endogenous fusel acid stress response [12].
The War1p protein belongs to the fungal-specific
Zn(II)
2
Cys
6
zinc finger family of transcriptional regula-
tors with some 54 other putative members in S. cere-
visiae [16]. These are implicated in various important
cellular processes, including amino-acid [17] and galac-
tose [18] metabolism, nitrogen source utilization [16],
peroxisomal proliferation [19,20], respiration [21,22]
and even pleiotropic drug resistance (PDR) [11]. For
example, Pdr1p and Pdr3p are key players in yeast
PDR development, because they control ABC drug

efflux pumps such as Pdr5p [23,24], Snq2p [25,26] and
Yor1p [27], all of which are involved in PDR [10,11].
Most regulators harbor a binuclear DNA-binding zinc
cluster at the N-terminus, whereas the acidic activation
domain is usually present at the C-terminus. The mid-
dle homology region (MHR) bridging the DNA-bind-
ing and the activation domain may control the activity
or specificity of the transcription factor, as deletions or
mutations in this region often result in constitutive
activity [22,28–30]. Notably, a WAR1 orthologue has
been identified in the human fungal pathogen Candida
albicans [31]. Consistent with its role in S. cerevisiae,
this WAR1 is also required for sorbate tolerance.
Moreover, our group recently identified the Candida
glabrata homologue of War1p (C. Gregori and
K. Kuchler, unpublished work). Preliminary experi-
ments show that it is also required for a response to
weak organic acids in the human fungal pathogen
C. glabrata, demonstrating the evolutionary conserva-
tion of this weak-acid stress in the fungal kingdom
(C. Gregori and K. Kuchler, unpublished work).
Secondly, we applied a classical genetic screen using
a PDR12prom-lacZ reporter gene to identify compo-
nents of the weak-acid response pathway. Here, we
report the results of the genetic approach, which leads
to the isolation of two war1 mutant alleles that are
unable to drive Pdr12p induction. Remarkably, the
genetic screen identified mutations in the WAR1 gene
only, indicating that weak-acid stress response requires
two major components, a dedicated stress regulator

and the Pdr12p efflux pump. The defective War1-42 p
mutant is no longer phosphorylated upon stress and
C. Gregori et al. Yeast weak organic acid stress adaptation
FEBS Journal 274 (2007) 3094–3107 ª 2007 The Authors Journal compilation ª 2007 FEBS 3095
unable to bind to cis-acting WARE motifs, suggesting
that activation of War1p or its binding to WARE is
tightly linked to its post-translational modification.
Results
Isolation of sorbate-sensitive mutant strains
To identify components of the stress response pathway
that mediates induction of the Pdr12p efflux pump, we
set up a classical mutagenesis screen. For the isolation
of mutant cells that fail to induce PDR12 upon weak
acid challenge, we constructed a reporter strain carrying
the lacZ gene driven by the PDR12 promoter integrated
into the ura3 loci of two different genetic backgrounds,
creating the strains YCS12ZI and YAK3. These strains
were grown to the exponential growth phase, plated
and irradiated with UV light to randomly introduce
mutations. After a 2-day incubation, colonies were
replica-plated on plates containing 5 mm sorbate and
the dye X-Gal to induce the PDR12 promoter and to
visualize LacZ expression. In a first round of screening,
we obtained 111 white colonies (62 for YAK3 and 49
for YCS12Z-I). To determine if the white color resulted
from a lack of PDR12 promoter induction, and thus no
expression of lacZ, these colonies were re-screened for
their Pdr12p protein concentrations by immuno-
blotting. Although several mutants showed reduced
Pdr12p concentrations under stress conditions (data not

shown), only two mutant strains, 42 and 28, lacked
detectable Pdr12p induction upon sorbate stress
(Fig. 1A). Both mutants were back-crossed with the
wild-type several times to clean up the genetic back-
ground and determine whether the phenotype was
caused by mutations in a single gene. As tetrad analysis
revealed a 2 : 2 cosegregation of sorbate sensitivity with
the inability of lacZ induction, both mutations must
reside in a single gene (data not shown). Growth-inhibi-
tion assays (Fig. 1B) showed that mutants 28 and 42
grew at a sorbate concentration of up to 0.25 mm. The
pdr12D control strain was viable, but exhibited reduced
growth on 0.5 mm sorbate plates, and failed to grow on
1mm sorbate, whereas the wild-type control even grew
at concentrations above 1 mm. Therefore, we isolated
two yeast mutants with defects in a single gene repre-
senting at least one component of the weak-acid
response machinery that acts through Pdr12p induction
to trigger adaptation.
Identification of mutated genes
In addition to the classical genetic approach, we
recently pursued a functional genomics approach to
identify regulators of weak organic acid resistance.
Making use of the EUROSCARF haploid deletion
strain collection of S. cerevisiae [14] (EUROSCARF,
Germany; http// />mikro/euroscarf/), we tested all viable transcription
factor deletion strains for their ability to grow in the
presence of sorbic acid. This approach identified the
transcription factor essential for Pdr12p induction and
hence weak-acid resistance, War1p [15]. To determine

if the mutants isolated in the classical genetic screen
are allelic to WAR1, appropriate selection markers
were integrated and the strains subjected to com-
plementation analysis. Figure 2 shows the growth
phenotypes of the resulting diploid strains on yeast pep-
tone ⁄ dextrose (YPD), pH 4.5, with different sorbate
A
B
WT war1-28 war1-42
pdr12Δ
WT
war1-28
war1-42
pdr12Δ
YPD pH 4.5
control
0.25 m
M
0.5 m
M
1 m
M
+ Sorbate
Pdr12p
control
sorbate
-+-+-+-+
Fig. 1. war1 mutants are sorbate-sensitive and fail to induce
Pdr12p upon sorbate stress. (A) The strains W303-1A (WT), YAK4
(War1-28), YCS42-D4 (War1-42) and control strain YBB14 (pdr12D)

were grown in YPD to an A
600
of  1. The cultures were split and
one half was stressed with 8 m
M sorbate for 1 h. Cell extracts
equivalent to 0.5 A
600
were separated by SDS ⁄ PAGE (7% gel), and
the immunoblots were decorated with polyclonal anti-Pdr12p
serum. A cross-reaction to the antibodies served as loading control.
(B) The strains W303-1A (WT), YAK4 (War1-28), YCS42-D4 (War1-
42) and YBB14 (pdr12D) were grown in YPD to an A
600
of  1.
Then the A
600
was adjusted to 0.2, and the cells were spotted
along with three 1 : 10 serial dilutions on YPD, pH 4.5, containing
the indicated sorbate concentrations. Growth was monitored after
a 48-h incubation at 30 °C.
Yeast weak organic acid stress adaptation C. Gregori et al.
3096 FEBS Journal 274 (2007) 3094–3107 ª 2007 The Authors Journal compilation ª 2007 FEBS
concentrations. Mutants 28 and 42, as well as the
war1D deletion, when in combination with a wild-type
gene, displayed similar growth to the diploid wild-type
strain. Thus, the mutant alleles, 28 and 42, are reces-
sive for sorbate growth, and a heterozygous wild-
type ⁄ war1D strain did not show any haplo-insufficiency
phenotypes (Fig. 2). However, when mutants 28 and
42 were crossed with the war1D strain, the diploid

strains remained hypersensitive to sorbate, and dis-
played the same growth behavior as the pdr12D
control strain. These data suggest that the mutants
isolated in the UV mutagenesis screen were allelic to
WAR1. Thus, the mutant alleles were named War1-28
and War1-42, respectively. Interestingly, diploid War1-
28 ⁄ War1-42 cells were more resistant than war1D ⁄
mutant diploids, suggesting a possible cross-comple-
mentation of mutant alleles, and implying that War1p
acts as a dimer [15].
Identification of the mutations in war1
To identify the actual mutations leading to the loss-of-
function phenotypes, the defective war1 alleles were
amplified by PCR from genomic DNA obtained from
the mutant strains and subjected to DNA sequencing.
Sequencing of both DNA strands of War1-28 identi-
fied an A to T mutation at position 1286, and a
change of C to T at position 1288, the latter introdu-
cing a translational stop codon (Fig. 3). At the amino-
acid level, these mutations resulted in a N429I residue
exchange, and the nonsense mutation leads to trun-
cated War1-28p protein (Fig. 3A). For the mutant
War1-42 allele, four clustered mutations were found:
deletion of A2286, T2287 and T2288, and the G2291T
transversion. These four mutations caused three amino
acid changes, namely K762N, F763M and the R764D
deletion. The rest of the protein remained unchanged.
As depicted in the cartoon (Fig. 3A), War1p is repre-
sentative of the binuclear Zn(II)
2

Cys
6
transcription
factor family, all members of which contain a DNA-
binding zinc finger at their N-terminus (amino acids
75–111), followed by two predicted nuclear localization
signals (NLS amino acids 106–123 and 286–303), and
a coiled-coil domain mediating protein–protein interac-
tions. The putative transcriptional activation domain is
located near the C-terminus, residues 911–937. Hence,
a loss-of-function phenotype of War1-28 may easily be
explained by the absence of the activation domain,
whereas the effect of the mutations in War1-42 on
War1p function is not immediately obvious.
Characterization of War1-42p and its
post-translational modification
To determine if the mutant proteins are properly
expressed, we epitope-tagged both War1-28p and
War1-42p at the C-terminus by genomic integration of
a triple 3HA epitope, creating the strains YBB30
and YBB31, respectively. Immunoblotting of protein
extracts from exponentially growing cultures revealed
that both mutant proteins displayed a mobility corres-
ponding to their predicted molecular mass (Fig. 3B).
However, when compared with the wild-type, the
steady-state concentrations of mutant War1-42p-3HA
appeared to be markedly reduced. Notably, the con-
centrations of the truncated War1-28p-3HA appeared
slightly increased, implying that the stability of the
protein is affected by the different mutations.

To address this point, we performed cycloheximide
chase experiments. The strains, YAK111, YBB30 and
YBB31, were grown in YPD to an A
600
of  1; then
cycloheximide was added to block protein synthesis,
and samples were collected at the indicated time points
for immunoblotting (Fig. 3C). The results show that
the wild-type protein was quite stable, with a half-life
of 100–120 min. Likewise, War1-28p-3HA was detect-
able throughout the whole chase period (Fig. 3C). In
contrast, War1-42p-3HA displayed a much faster pro-
teolytic turnover, as it was already below the detection
limit 40 min after cycloheximide addition (Fig. 3C).
Thus, the low steady-state concentrations of War1-
42p-3HA may be explained by its reduced stability.
Notably, sorbate failed to influence War1p stability, as
the half-life was unchanged under stress (data not
shown).
WT/WT
WT/war1Δ
WT/28
WT/42
war1Δ/28
war1Δ/42
28/42
pdr12Δ
YPD pH 4.5
control
0.5 m

M 1 mM
+ Sorbate
Fig. 2. The sorbate-sensitive mutants carry loss-of-function alleles
of WAR1. The strains W303-D (WT ⁄ WT), YBB21 (WT ⁄ war1D),
YBB24 (WT ⁄ 28), YBB22 (WT ⁄ 42), YBB25 (war1D ⁄ 28), YBB26
(war1D ⁄ 42), YBB23 (28 ⁄ 42) and YBB14 (pdr12D) were grown to an
A
600
of 1, diluted to A
600
of 0.2 and spotted on to YPD, pH 4.5,
agar plates containing the indicated sorbate concentrations along
with three 1 : 10 serial dilutions. Colony growth was inspected
after 48 h at 30 °C.
C. Gregori et al. Yeast weak organic acid stress adaptation
FEBS Journal 274 (2007) 3094–3107 ª 2007 The Authors Journal compilation ª 2007 FEBS 3097
Stress-induced phosphorylation is absent
in War1-42p
Whereas for War1-28p the inability to induce PDR12
transcription was attributable to the lack of the activa-
tion domain, the explanation was not obvious for
War1-42p. Notably, we have previously shown that
PDR12 induction coincides with War1p phosphoryla-
tion [15]. Therefore, we determined the post-transla-
tional modification status of War1-42p-3HA under
both stressed and nonstressed conditions using immu-
noblotting (Fig. 3D). The cultures were grown to an
A
600
of  1, split, and one half was treated with 8 mm

sorbate. After 30 min, cells were harvested, and protein
extracts prepared and subjected to immunoblotting. As
reported previously [15], wild-type War1p migrated as
a double band in unstressed cells and shifted to slower
mobility forms upon sorbate addition (Fig. 3D). In
contrast, no mobility shift was detectable for War1-
42p-3HA, as it migrated as a single band under both
stressed and nonstressed conditions (Fig. 3D). There-
fore, the post-translational modification pattern of
War1p, which is intimately linked to PDR12 stress
induction, is absent in the War1-42p-3HA mutant, indi-
cating that phosphorylation may be an essential step
in War1p activation. Remarkably, War1-42p-3HA
from unstressed cells exhibited a faster mobility on
SDS ⁄ polyacrylamide gels than authentic War1p
(Fig. 3D), suggesting that the basal modification in the
absence of stress was also affected in War1-42p-3HA.
Functional analysis of single-residue changes,
K762N, F763M and R764D
The War1-42 allele contains a cluster of four muta-
tions, leading to three residue changes. To address
which mutation alone or in combination with another
one causes the phenotype, we constructed the CEN-
based plasmids pCGWAR1-K762N, pCGWAR1-
F763M and pCGWAR1-R764D carrying the single
mutations, respectively. Each of the three plasmids
expressed a mutated version War1p with only one of
three residue changes of War1-42p. To determine the
phosphorylation status of War1p-K762N, War1p-
war1-28

N429I
STOP
war1-42
K762N
F763M
R764Δ
Zn
NLS
AD
NLS
A
WT-3HA
28-3HA
42-3HA
WT
loading control
cross reaction
B
War1p-3HA
War1-28p-3HA
War1p-3HA
War1-28p-3HA
War1-42p-3HA
0 20 40 60 80 100 120 min CHX
C
WT-3HA
WT-3HA
42-3HA
42-3HA
WT

WT
D
unstressed + 8 mM sorbate
30 min
War1p-3HA
Fig. 3. Organization, expression, stability and modification of War1p
variants. (A) The cartoon depicts the localization of mutations in the
WAR1 gene abrogating its function as a specific Pdr12p regulator.
Zn, zinc finger; AD, activation domain. Cartoon not drawn to scale.
(B) Expression and stability of the wild-type and mutant War1p vari-
ants. Cultures of the strains YAK111 (WT-3HA), YBB31 (28-3HA),
YBB30 (42-3HA) and YPH499 (WT) were grown in YPD to an A
600
of  1 and harvested. Yeast crude protein extracts equivalent to 1
A
600
were separated by PAGE (10% gel) and analyzed by immuno-
blotting using the 12CA5 HA antibody. Cross-reactions to the HA
antibody served as a loading control. (C) The strains YAK111
(War1p-3HA), YBB31 (War1-28p-3HA) and YBB30 (War1-42p-3HA)
were grown in YPD to an A
600
of  1, then cycloheximide (CHX)
was added to a final concentration of 0.1 mgÆmL
)1
, and samples
were taken at the indicated time points. Extracts (0.5 A
600
for
War1p-3HA and War1-28p-3HA, 1.5 A

600
for War1-42p-3HA) were
fractionated by SDS ⁄ PAGE (10% gel), followed by immunodetec-
tion of the War1p-3HA and variants by monoclonal 12CA5 HA anti-
body. (D) The strains YAK111 (WT-3HA), YBB30 (42-3HA) and
YPH499 (WT) were grown in YPD to an A
600
of  1, then the cul-
tures were split, and one half was treated with 8 m
M sorbate for
30 min. Crude cell extracts (equivalent to 1.5 A
600
for 42-HA and
0.5 A
600
for WT-HA and WT) were separated by SDS ⁄ PAGE (7%
gel) and immunodetected using the monoclonal 12CA5 HA anti-
body.
Yeast weak organic acid stress adaptation C. Gregori et al.
3098 FEBS Journal 274 (2007) 3094–3107 ª 2007 The Authors Journal compilation ª 2007 FEBS
F763M and War1p-R764D, as well as their ability to
induce Pdr12p expression, we performed immuno-
blotting under stressed and nonstressed conditions
(Fig. 4A). Cultures of the war1D strain YAK120,
harboring pCGWAR1-K762N, pCGWAR1-F763M,
pCGWAR1-R764D or the control plasmid expressing
wild-type WAR1, were grown to an A
600
of  1. The
cultures were split in half; one half was treated with

8mm sorbate for 30 min, and the other left unstressed.
Cells were then harvested, and protein extracts were
prepared and subjected to immunoblotting. Whereas
War1p-F763M behaved as the wild-type War1p under
unstressed conditions, War1p-K762N showed a slightly
different modification pattern in unstressed cells. In
contrast with wild-type War1p, which migrated as a
double band under nonstressed conditions, the lower-
migrating band was hardly detectable in War1p-K762N
(Fig. 4A). However, sorbate shifted both War1p-
K762N and War1p-F763M to a slower mobility, as
was also observed for the wild-type control. In con-
trast, War1p-R764D remained unmodified in response
to sorbate stress. The mobility shift in response to
stress is tightly linked to PDR12 induction, which was
absent in the strain expressing War1p-R764D (Fig. 4A).
In contrast, strains expressing the mutants War1p-
K762N and War1p-F763M showed greatly increased
Pdr12p concentrations in both the absence and pres-
ence of weak-acid stress. Addition of sorbate did not
further increase Pdr12p expression levels (Fig. 4A),
demonstrating that the War1p-K762N and War1p-
F763M single mutants are gain-of-function variants.
However, their hyperactivity is suppressed by the pres-
ence of the additional R764D deletion in War1-42p.
Furthermore, we tested the strains expressing the
mutant War1p variants for their ability to grow on
YPD, pH 4.5, in the presence or absence of different
sorbate concentrations (Fig. 4B). Consistent with the
immunoblotting data, the War1p-R764D strain was as

hypersensitive to sorbate as the war1D control because
of the inability to induce Pdr12p, whereas War1p-
K762N and War1p-F763M showed normal growth
when compared with wild-type cells (Fig. 4B). Nota-
bly, the hyperactivity of War1p-K762N and War1p-
F763M variants failed to cause hyper-resistance to
weak organic acids (data not shown).
Because War1-42p displayed reduced protein stabil-
ity (Fig. 3C), cycloheximide chase experiments were
also performed with the War1p-R764D single mutant
exactly as described above for War1-42 p (Fig. 3C).
Detection of the different War1p variants demonstra-
ted a markedly reduced stability of War1p-R764D
compared with wild-type, although protein concentra-
tions did not decrease as fast as for War1-42p
F763MWT K762N
R764Δ
-
+
-
+
-
+
-
+
8 m
M sorbate
War1p
Pdr12p
Pdr5p

WT
war1Δ
K762N
F763M
R764
Δ
YPD pH 4.5
+ 0.5 m
M sorbate + 1 mM sorbate + 2 mM sorbate
0 20 40 60 80 100 120 min CHX
War1p
War1p-R764Δ
War1-42p
B
A
C
Fig. 4. War1p-764D causes sorbate hypersensitivity and fails to
induce Pdr12p. (A) Cultures of YAK120 cells were transformed with
pCGWAR1, pCGWAR1-K762N, pCGWAR1-F763M or pCGWAR1-
R764D and grown in YPD to an A
600
of  1; cultures were split in
half, and one half was treated with 8 m
M sorbate for 30 min, while
the other remained untreated. Crude cell extracts (0.5 A
600
) were
separated by SDS ⁄ PAGE (7% gel), and immunodetected using
polyclonal antibodies against War1p, Pdr12p and Pdr5p. (B) YAK120
was transformed with pCGWAR1, pCGWAR1-K762N, pCGWAR1-

F763M or pCGWAR1-R764D and grown to an A
600
of 1, diluted to
A
600
¼ 0.2, 0.02 and 0.002; culture aliquots were spotted on to
YPD, pH 4.5, agar plates containing sorbate concentrations as indi-
cated. Colony growth was inspected after 48 h at 30 °C. (C) The
strain YAK120 transformed with pCGWAR1 or pCGWAR1-R764D
and strain YCS42-D4 (War1-42p) were grown in YPD to an A
600
of
 1. Cycloheximide (CHX) was added at a final concentration of
0.1 mgÆmL
)1
; samples were taken at the indicated time points.
Cell-free extracts (0.5 A
600
for War1p, 1.0 A
600
for War1-42p and
War1-R764Dp) were separated by SDS ⁄ PAGE (7% gel) and trans-
ferred to nitrocellulose membranes. War1p and variants were
detected by immunoblotting using polyclonal antibodies to War1p.
C. Gregori et al. Yeast weak organic acid stress adaptation
FEBS Journal 274 (2007) 3094–3107 ª 2007 The Authors Journal compilation ª 2007 FEBS 3099
(Fig. 4C). This may indicate that the cluster of three
residue changes in War1-42p, as well as single residue
changes, even if two of them lead to a gain-of-func-
tion, change War1p folding, thereby destabilizing

War1p more than the loss of a single amino acid as in
War1-R764Dp.
War1-42p localizes to the nucleus but is unable
to bind to the WARE in vivo
We have previously demonstrated that War1p is a nuc-
lear protein [15]. Although the mutational changes in
War1-42p left the DNA-binding domain and both NLS
unaffected, we wanted to test whether War1-42p is also
properly localized to the nucleus. Hence, we carried out
fractionation experiments using purified nuclear frac-
tions from various strains. Subcellular fractions were
isolated by following a gentle cell lysis procedure to
preserve nuclear integrity, and subjected to immuno-
blotting using polyclonal antibodies specific for War1p,
the nuclear marker protein Swi6p and the cytoplasmic
hexokinase Hxk1p (Fig. 5A). As shown in Fig. 5A,
War1-42p, like the wild-type control War1p, localized
to the nucleus in the steady state. As expected, Hxk1p
was predominantly found in the soluble cytoplasmic
fraction. The signal for Hxk1p in the nuclear fraction is
due to normal unavoidable contamination of the nuc-
lear fraction with cytosolic proteins. However, the nuc-
lear marker, Swi6p, entirely cofractionated with both
War1-42p and wild-type War1p (Fig. 5A), demonstra-
ting that the normal nuclear localization of War1-42p
is unaffected by the mutations. No immunoreactive
material was detectable in war1D cells, confirming the
specificity of the polyclonal anti-War1p serum. Nota-
bly, the polyclonal antibodies also detected a War1p
degradation product (Fig. 5A), which was not recog-

nized by the monoclonal HA antibody (Fig. 3). Hence,
the lack of function in War1-42p is probably a conse-
quence of impaired activation or direct binding to the
WARE rather than aberrant cellular localization.
War1-42 strains cannot tolerate sorbate, suggesting
that they may lack the capacity for proper modification
of War1p under stress conditions. To exclude this pos-
sibility, we introduced War1-42p into a wild-type back-
ground to obtain strain YBB32. With this strain
carrying one wild-type and one mutated allele of
WAR1, we repeated the stress experiments described
above and checked for the mobility of War1-42pby
immunoblot analysis. Even in the presence of wild-type
War1p and a normal stress response, War1-42p
remains unmodified (data not shown), suggesting that
the mutations prevent normal modification of the
transcription factor rather than indirect effects, which
321
1+2 1+3 2+3
WT
war1-42
war1Δ
WT
war1-42
war1Δ
*
*
42
WT
war1Δ

CNSCSCNSN
War1p
Swi6p
Hxk1p
B
A
Fig. 5. Nuclear War1-42p is unable to decorate the PDR12 promoter
in vivo. (A) Subcellular fractions were prepared from wild-type cells
(WT), War1-42 mutants (42) and cells lacking War1p (war1D) as des-
cribed in Experimental procedures. About 2 A
600
equivalents were
subjected to immunoblotting using polyclonal antibodies against
War1p, Swi6p and Hxk1p. S, Total input; N, nuclear pellet; C, cyto-
plasmic fraction. (B) YPH499 wild-type (WT), YAK110 (war1D) and
YCS42-D4 (War1-42) cells were grown to the early exponential
growth phase and treated with dimethyl sulfoxide to methylate
DNA. For in vivo footprinting, chromosomal DNA was prepared and
used as a template for primer extension with a labeled oligonucleo-
tide primer corresponding to )497 to )472 of the PDR12 promoter.
The reaction mixture was resolved through a sequencing gel,
exposed to a phosphoimaging screen, and signals were quantified.
Intensities of traces are compared in the indicated combinations. An
asterisk marks a protected G ()631) in the War1-42 allele. Differ-
ences are indicated by bars for deprotected G residues ()643,
)642, )617, )618) and aligned with the sequence of the region.
Yeast weak organic acid stress adaptation C. Gregori et al.
3100 FEBS Journal 274 (2007) 3094–3107 ª 2007 The Authors Journal compilation ª 2007 FEBS
would be a consequence of the impaired growth of
War1-42 mutants under stress.

Because nuclear localization of War1-42p was unaf-
fected, we determined whether War1-42p still binds to
the WARE in the PDR12 promoter. To clarify this,
we performed in vivo footprinting experiments in cells
expressing War1-42p. The strains YPH499 (wild-type),
YAK110 (war1D) and YCS42-D4 (War1-42) were cul-
tivated to the exponential growth phase and then trea-
ted with dimethyl sulfoxide to methylate guanines and
to some extent adenines in the genomic DNA. Chro-
mosomal DNA was isolated, and the methylation
status was determined by primer extension analysis
(Fig. 5B). Comparison of the methylation patterns of
the different strains revealed that the mutant War1-42
and war1D exhibit almost identical patterns (Fig. 5B,
2+3). In contrast, when both were compared with the
wild-type (Fig. 5B, 1+2 and 1+3), deprotection at the
same nucleotides (black bars) was observed. These
results clearly indicate that War1-42p is unable to dec-
orate the WARE in the PDR12 promoter in vivo,
nicely explaining the loss-of-function phenotype and
the sorbate hypersensitivity of War1-42 mutant cells.
Taken together, our genetic screen identified two muta-
tions in the putative MHR region of the War1p tran-
scriptional regulator, suggesting that the MHR is
essential for War1p function. Mutations in the MHR
may cause structural changes that impair post-transla-
tional phosphorylation, affecting DNA binding of
War1p or the recruitment of other as yet unknown
coregulators.
Discussion

We are interested in dissecting the response pathway
necessary for cellular adaptation to stress from weak
organic acids in the yeast, S. cerevisiae. Using a func-
tional genomic approach, we have identified the
War1p regulator as the dedicated transcription factor
required for Pdr12p induction following weak-acid
stress exposure [15]. In this study, we report the iso-
lation and characterization of two loss-of-function
war1 alleles that give rise to War1p variants that are
unable to mediate Pdr12p induction in response to
sorbate stress. We exploited a classical genetic screen,
taking advantage of a lacZ reporter driven by the
PDR12 promoter, which otherwise controls expression
of the Pdr12p weak-acid anion-efflux pump [7]. After
UV mutagenesis, we screened more than 10 genome
equivalents of mutant colonies for their capacity in
PDR12 induction. We expected to isolate mutants in
membrane sensors, signaling components such as
kinases, phosphatases and perhaps transcriptional
regulators. However, most remarkably, only two yeast
mutants were isolated in which sorbate-mediated
induction of Pdr12p was completely abolished
(Fig. 1A). The weak-acid hypersensitivity of both
mutant strains was attributed to mutations in the
WAR1 gene (Fig. 3). DNA sequencing identified the
mutations in the nonfunctional war1 alleles. Whereas
War1-28 encoded a truncated regulator which was due
to a stop codon, War1-42 carried three residue chan-
ges close to the C-terminus. Hence, the genetic
approach yielded only mutant variants of the War1p

regulator, suggesting that War1p is the major and per-
haps only stress regulator of Pdr12p.
The genetic approach confirms our strategy of
using functional genomics which led to the identifica-
tion of War1p. Nevertheless the genetics data allow
several interpretations about the function of War1p.
Firstly, none of the signaling components except
War1p, including the War1p kinase(s), appear to be
essential, perhaps because of redundant functions in
the pathway. Secondly, Pdr12p and War1p represent
the key elements of the response pathway and there
are no other essential components. Thirdly, we can-
not entirely exclude the possibility that other
mutants have been missed by the low amount of
sorbate used for the PDR12 promoter induction
screen because of weak-acid hypersensitivity. How-
ever, as Pdr12p is the major determinant of weak-
acid resistance, and the main target of War1p [32],
we reason that defects in genes encoding components
acting upstream of War1p should not display higher
sorbate sensitivities than nonfunctional War1p vari-
ants themselves.
The sorbate hypersensitivity phenotype of War1-28
cells can easily be explained, because this allele carries
a nonsense mutation leading to a truncated War1-28p
(Fig. 3A). Hence War1-28p lacks the C-terminal acti-
vation domain, which is necessary for transcriptional
activation of the target genes by other Zn(II)
2
Cys

6
transcription factors [33]. Notably, the truncated
War1-28p protein, although expressed at higher levels
than the wild-type, does not interfere with the function
of authentic War1p in diploid cells (Fig. 2), which
might be a direct consequence of impaired dimer for-
mation of War1p or a lack of DNA binding. Indeed,
in the case of War1-42p, in vivo footprinting data
(Fig. 5) indicate an inability to bind to the WARE,
which is normally decorated by wild-type War1p in the
presence or absence of the stress agent [15]. Thus, the
lack of PDR12 stress induction by War1-42p is per-
haps due to its inability to bind to the promoter
WARE of its target gene. Alternatively, the mutations
may also reduce the binding affinity of the War1-42p,
C. Gregori et al. Yeast weak organic acid stress adaptation
FEBS Journal 274 (2007) 3094–3107 ª 2007 The Authors Journal compilation ª 2007 FEBS 3101
thereby causing impaired assembly of the active tran-
scription complex.
A lack of DNA binding by War1-42p may be
explained in several ways. First, immunoblotting
cycloheximide chase experiments indicated that War1-
42p shows lower steady-state protein concentrations
and decreased stability compared with the wild-type
War1p (Fig. 4). Hence, the amount of protein might
simply be too low to allow binding to the target DNA.
Secondly, the mutations may reduce affinity for
WARE binding or prevent the formation of War1p
dimers, which appears necessary for War1p function
[15]. Thirdly, reduced protein concentrations may still

allow WARE binding, but other determinants are pre-
venting the DNA recognition. Although band shift
experiments using glutathione S-transferase (GST)-
War1p suggested that WARE binding does not require
additional factors or modifications [15], this may not
entirely reflect the situation in vivo. As shown for
Gal4p, the prototype zinc cluster transcription factor,
its DNA-binding properties can be different in vitro
and in vivo and may involve sequences in the regula-
tory domains distinct from the zinc finger or even
additional factors [34]. Therefore, the war1 mutations
identified perhaps destroy structural features necessary
for the interaction with accessory proteins involved in
target site recognition, WARE binding or post-transla-
tional modification of War1p.
Genetic separation of the clustered mutations in
War1-42p revealed that R764D displays a complete
loss-of-function phenotype. In sharp contrast, how-
ever, the residue changes, K762N and F763M, lead to
constitutive War1p hyperactivity (Fig. 4A). Strikingly,
War1p-R764D is no longer phosphorylated in response
to sorbate challenge. Conversely, War1p-K762N and
War1p-F763M are phosphorylated upon stress,
although they are already active in the absence of
sorbate stress (Fig. 4A). Therefore, in constitutively
active War1p variants, sorbate stress is tightly linked
to phosphorylation, even if the protein is already acti-
vated. War1-42p is functionally inactive and also not
phosphorylated in response to weak-acid treatment
(Fig. 3D). Hence, the R764D mutation can be consid-

ered dominant for War1p loss-of-function when pre-
sent in combination with the hyper-activating
mutations, K762N and F763M.
The basal modification status of War1-42p is differ-
ent from wild-type War1p, as they display distinct
mobilities on immunoblotting. DNA binding in vivo
may well require basal post-translational modifications,
as present in wild-type War1p but absent in the
mutant variant (Fig. 3D). These modifications either
directly influence the DNA-binding capability or are
necessary for the interaction with another as yet
unknown cofactor that would facilitate binding to the
PDR12 promoter.
The fact that loss-of- function mutations reside out-
side the zinc finger or the NLS suggests an altered
conformation or structure. This is consistent with the
apparent absence of stress-induced phosphorylation in
War1p-R764D and its reduced protein stability. Thus,
only massive folding changes can explain the inactiv-
ity of War1p-R764D, which may hinder phosphoryla-
tion. Mutations may also affect the structure of
confined domains such as the MHR rather than the
whole tertiary structure. Hence, the lack of phos-
phorylation is most likely a consequence of massively
altered War1p conformation rather than altered struc-
ture of the kinase targets themselves. Further, the
residue changes in War1-42p do not involve serine,
threonine, tyrosine or histidine (Fig. 3A). In any case,
the nonphosphorylated war1 alleles are nonfunctional,
indicating that certain post-translational modifications

are essential for War1p to induce transcription of
PDR12 upon stress.
From the homologies between zinc finger regulators,
a defined MHR [33] is not immediately apparent in
War1p. However, it seems plausible that the stretch
carrying the residue changes is functionally similar to
the MHR. Because the War1p-K762N and War1p-
F763M mutants in this stretch are constitutively active,
the MHR of War1p is likely to play a major role in
the regulation of its transcriptional activity, as well as
in the specificity of target site recognition. If the regu-
lator is present in limiting amounts, a reduced or
altered specificity because of lack of a functional
MHR will remove the protein from its binding sites
in vivo [33]. However, we wish to provide another
explanation for the loss-of-function in War1-42p. The
drastic decrease in the total amount of the transcrip-
tional regulator in combination with a reduction in
sequence recognition specificity may account for the
observed lack of WARE binding by War1-42p. Defect-
ive nuclear localization can be excluded, because the
NLSs are not affected by the mutations, and because
the mutant proteins display proper nuclear localiza-
tion. Another possibility is that War1-42p binds to
WARE with much lower affinity, insufficient to be
detected by in vivo footprinting (Fig. 5B).
Thus, more information about the structure and
potential interaction partners of War1p is required. For
instance, the War1-42 mutant can be used as a tool to
identify intragenic suppressor mutations. Furthermore,

high-copy or second-site suppressors may lead to the
identification of unknown War1p-interacting partners.
This seems to be a feasible and promising approach
Yeast weak organic acid stress adaptation C. Gregori et al.
3102 FEBS Journal 274 (2007) 3094–3107 ª 2007 The Authors Journal compilation ª 2007 FEBS
considering the suppression of the constitutive activity
of K762N and F763M by an additional R764D dele-
tion. Finally, the War1-42 mutant allele should be use-
ful as a tool for learning more about the molecular
structure, as well as the protein–protein and protein–
DNA interactions, of this binuclear zinc transcriptional
regulator. The polyclonal antibodies to War1p should
be useful in identifying the upstream components of the
response pathway, including War1p-specific kinases
and phosphatases implicated in the modulation of
War1p activity during adaptation to stress induced by
weak organic acids.
Experimental procedures
Yeast strains, growth conditions, and growth
inhibition assays
Rich medium (YPD) and synthetic medium were prepared
essentially as described elsewhere [35]. Unless otherwise
indicated, all yeast strains were grown routinely at 30 °C.
S. cerevisiae strains used in this study are listed in Table 1.
To determine weak-acid susceptibility, exponentially grow-
ing cultures were adjusted to A
600
of 0.2 and diluted 1 : 10,
1 : 100 and 1 : 1000. Equal volumes of these serial dilutions
were spotted on to YPD, pH 4.5, plates containing the indi-

cated sorbate concentrations exactly as previously described
[15].
Gene disruptions and strain constructions
The deletion of WAR1 was performed by a PCR-based
method using the disruption cassette of the plasmid pFA6a-
HIS3MX6 [36]. The PDR12 gene was disrupted with a
hisG-URA3-hisG cassette from the plasmid pYM63 [7].
For epitope tagging of the wild-type or mutant versions of
WAR1, the triple HA tag or GFP tag was amplified from
plasmids pFA6a-3HA-KANMX6 or pFA6a-3HA-HIS3MX6
[37] using appropriate primers, followed by integration at the
genomic locus.
For the PDR12prom-lacZ reporter construct used in the
UV-mutagenesis screen, we amplified the PDR12 promoter
by PCR, introducing an EcoRV site at position )1168 and
a HindIII site at position +8. The EcoRV–HindIII frag-
ment was then cloned into the vector YIp357 [38], yielding
plasmid pCS12Z-I. The correct sequence of the insert was
verified by DNA sequencing. To construct the strain carry-
ing lacZ under the control of the PDR12 promoter, plasmid
Table 1. Yeast strains used in this study.
Strains Genotype Source
W303–1B MATa ura3-1 leu2-3112 his3-11,15 trp1-1 ade2-1 can1-100
(W303-1A - MATa, W303-D - MATa ⁄ a)
[45]
YPH499 MATa ura3-52 leu2-D1 his3-D200 trp1-D1 ade2-10
oc
lys2-801a [39]
YCS12ZI MATa ura3-52::pCS12ZI URA3 (isogenic to YPH499) This study
YAK2 MATa ura3-1::pCS12ZI URA3 (isogenic to W303-1B) [15]

YAK3 MATa ura3-1::pCS12ZI URA3 LEU2 (isogenic to W303-1B) [15]
YAK4 MATa ura3-1::pCS12ZI URA3 LEU2 War1-28 (isogenic to W303-1B) [15]
YAK110 MATa war1D::HIS3MX6 (isogenic to YPH499) This study
YAK111 MATa WAR1-3HA KANMX6 (isogenic to YPH499) [15]
YAK120 MATa war1D::HIS3MX6 (isogenic to W303-1A) [15]
YCS42-D4 MATa ura3-52::pCS12ZI URA3 War1-42 (isogenic to YPH499) This study
YBB14 MATa pdr12::hisG-URA3-hisG (isogenic to W303-1A) This study
YBB20 MATa ⁄ a LEU2 war1D::HIS3MX6 (isogenic to W303-D) This study
YBB22 MATa ⁄ a ura3-1 leu2-3112 trp1-1 ade2-1 can1-100 ura3-52::pCS12ZI
URA3 leu2-D1 his3-D200 trp1-
D1 ade2-10
oc
lys2-801
a
War1-42
This study
YBB23 MATa ⁄ a ura3-1::pCS12ZI URA3 LEU2 his3-11,15 trp1-1 ade2-1 can1-100
War1-28 ura3-52::pCS12ZI URA3 leu2-D HIS3 trp1-D1 ade2-10
oc
lys2-801
a
War1-42
This study
YBB24 MATa ⁄ a ura3-1::pCS12ZI URA3 LEU2 HIS3 trp1-1 ade2-1 can1-100 War1-28
(isogenic to W303-D)
This study
YBB25 MATa ⁄ a ura3-1::pCS12ZI URA3 LEU2 his3-11,15 trp1-1 ade2-1 can1-100 War1-28
war1D::HIS3MX6 (isogenic to W303-D)
This study
YBB26 MATa ⁄ a ura3-1 leu2-3112 his3-11,15 trp1-1 ade2-1 can1-100 war1D::HIS3MX6 This study

ura3-52::pCS12ZI URA3 leu2-D1 his3-D200 trp1-D1 ade2-10
oc
lys2-801a War1-42
YBB30 MATa ura3-52::pCS12ZI URA3 War1-42-3HA HIS3MX6 (isogenic to YPH499) This study
YBB31 MATa ura3-1::pCS12ZI URA3 LEU2 War1-28-3HA HIS3MX6 (isogenic to W303-1B) This study
YBB32 MATa ⁄ a ura3-1 LEU2 his3-11,15 trp1-1 ade2-1 can1-100 ura3-52::pCS12ZI URA3
leu2-D1 his3-D200 trp1-D1 ade2-10
oc
lys2-801a War1-42-3-HAHIS3MX6
This study
C. Gregori et al. Yeast weak organic acid stress adaptation
FEBS Journal 274 (2007) 3094–3107 ª 2007 The Authors Journal compilation ª 2007 FEBS 3103
pCS12Z-I was linearized with StuI and integrated into the
ura3-52 locus of YPH499 or ura3-1 of W303-1B, creating
the strains YCS12Z-I and YAK2, respectively. Correct
genomic integration was confirmed by PCR, and lacZ func-
tionality was tested by b-galactosidase assays. For back-
crossing of war1 mutants after the screening, as well as
complementation analysis, additional markers were integra-
ted into the strains (Table 1), and diploids were selected by
double selection using appropriate auxotrophic markers.
UV mutagenesis
For mutagenesis by UV irradiation, the strains YCS12Z-I
or YAK3 were grown to the exponential growth phase, dilu-
ted, and plated on YPD at about 1000 cells ⁄ plate. Then, the
cells were treated in a UV Stratalinker 2400 (Stratagene, La
Jolla, CA) with a dose allowing for 80% survival of cells.
After 2 days incubation at 30 °C, the colonies were replica-
plated on X-Gal (5-chloro-4-bromo-3-indolyl-b-d -galacto-
side) plates containing 5 mm sorbate and grown for another

2 days. For YAK3, about 80 000 independent yeast colon-
ies, and for YCS12Z-I about 50 000 colonies, were screened
for the loss of sorbate-mediated lacZ induction. White col-
onies were re-streaked to single colonies on X-Gal plates
containing 5 mm sorbate for easier inspection of color devel-
opment. All colonies that remained white were re-screened
for Pdr12p concentrations by immunoblotting using poly-
clonal antibodies to Pdr12p [7]. Isolated mutant cells were
back-crossed at least three times to clean up the genetic
background and to verify 2 : 2 segregation of sorbate sensi-
tivity.
Site-directed mutagenesis and plasmid
construction
To generate the single-residue changes K762N, F763M and
R764D, a PCR fragment containing 170 bp of the WAR1
promoter and 2568 bp of the WAR1 ORF was ligated to
the vector pGEMT-easy (Promega, Mannheim, Germany)
resulting in the plasmid pIF1. This plasmid served as a tem-
plate for site-directed mutagenesis reactions using the
QuickChange Site-Directed Mutagenesis kit (Stratagene).
Site-directed mutagenesis was carried out exactly as recom-
mended by the manufacturer using the customized oligonu-
cleotide primers listed in Table 2. Mutations in the WAR1
sequence are indicated in bold italic letters in the primer
sequence. The plasmids obtained were named pIF1-762,
pIF1-763 and pIF-764, and successful mutagenesis was veri-
fied by DNA sequencing.
Fragments containing the indicated mutations were
cloned into a yeast vector as follows. A 4.32-kb PCR frag-
ment containing the entire WAR1 ORF, as well as 1 kb of

the 5¢-region and 0.45 kb of the 3¢-region were isolated by
PCR using genomic DNA from W303-1A. After digestion
with SalI and SacI, the PCR fragment was cloned into the
corresponding sites of pRS315 [39], resulting in the plasmid
pCGWAR1. The plasmid was sequenced to exclude PCR
errors. Plasmids pIF1-762, pIF1-763 and pIF-764 were
digested with NsiI, and the resulting 2440-bp fragments con-
taining the desired WAR1 mutations were used to replace
the corresponding wild-type fragment in pCGWAR1, yield-
ing the plasmids pCGWAR1-K762N, pCGWAR1-F763M
and pCGWAR1-R764D.
Production of rabbit polyclonal antibodies and
immunoblotting
Polyclonal antibodies to War1p were raised in rabbits
against a GST-War1p fusion protein containing 294 amino
acids of the C-terminal part of War1p (amino acids 650–
944) fused in-frame to the C-terminus of GST. The gene
fusion was constructed as follows. A 900-bp fragment of
WAR1 was generated by PCR, using the customized prim-
ers WAR1-GSTs (5¢-AAGAATTCTCCATGGGGGAAAT
GTCGCATACCATA-3¢) and WAR1-GSTas (5¢-CTGCA
GTCAAATGTCGACATTCATGAAAAGGTCTGTCC-3¢),
as described elsewhere [40]. The PCR product was digested
with EcoRI and SalI and cloned into the corresponding
sites of pGEX-5X-1 (Amersham Biosciences, Piscataway,
NJ). The resulting plasmid, pCG1950, allowed expression
of the C-terminal 294 amino acids of WAR1 fused in-frame
to the C-terminus of GST. Escherichia coli strain DH5a
harboring pCG1950 was grown on 37 °CtoanA
600

of 0.5.
Expression of the GST-War1p fusion protein was induced
at 30 °C for 2 h by adding isopropyl b-d-thiogalacto-
pyranoside to a final concentration of 0.2 mm. The
Table 2. Oligonucleotides for site-directed mutagenesis of WAR1. Bases leading to residue changes in mutagenic oligonucleotides are given
in bold italic letters.
Name Oligonucleotide sequence Source
K762Ns 5¢-CCCTTCAACAACTCTCTTTAC
AAC TTTAGGTATGTTATTGCG-3¢ This study
K762Nas 5¢-CGCAATAACATACCTAAA
GTT GTAAAGAGAGTTGTTGAAGGG-3¢ This study
F763Ms 5¢-CTTCAACAACTCTCTTTACAAA
ATG AGGTATGTTATTGCGTTATTTTG-3¢ This study
F763Mas 5¢-CAAAATAACGCAATAACATACCT
CAT TTTGTAAAGAGAGTTGTTGAAG-3¢ This study
R764Ds5¢-CCCTTCAACAACTCTCTTTACAAATTTTATGTTATTGCGTTATTTTGTC-3¢ This study
R764Das 5¢-GACAAAATAACGCAATAACATAAAATTTGTAAAGAGAGTTGTTGAAGGG-3¢ This study
Yeast weak organic acid stress adaptation C. Gregori et al.
3104 FEBS Journal 274 (2007) 3094–3107 ª 2007 The Authors Journal compilation ª 2007 FEBS
GST-War1p fusion protein was purified as described else-
where [15], except that the binding to the glutathione–Seph-
arose beads (Amersham) was performed at 4 °C for 16 h
and elution was carried out with 0.05% SDS. Removal of
SDS and concentration of the GST-War1p fusion protein
was carried out in a Centricon YM-10 centrifugal filter
device (Millipore, Billerica, MA). Antiserum to the purified
GST-War1p fusion protein was raised in rabbits using a
standard immunization regimen as described elsewhere [41].
Preparation of yeast cell extracts and immunoblotting
were performed exactly as described previously [41]. To

determine Pdr12p induction, strains were grown in YPD to
an A
600
of  1. The cultures were split, and one half was
stressed with 8 mm potassium sorbate for 1 h, and the
other left untreated as a control. Cell-free extracts equival-
ent to 0.5 A
600
)2 A
600
were separated by SDS ⁄ PAGE
(7% gel), followed by immunoblotting using polyclonal
antibodies to Pdr12p [7], War1p, Swi6p (a gift from
K. Nasmyth, IMP, Vienna, Austria) and Hxk1p (Biotrend,
Ko
¨
ln, Germany). The protein concentrations of wild-type
and mutant War1p variants were analyzed by immunoblot-
ting of extracts from exponentially growing yeast cultures
using the monoclonal HA antibody, 12CA5, or polyclonal
War1p antibodies using the ECL chemiluminescence detec-
tion and conditions as suggested by the manufacturer
(Amersham Biosciences).
Subcellular fractionation and analysis of protein
modification and stability
Subcellular fractionation experiments were performed essen-
tially as previously described [42]. Briefly, cells grown in
500 mL YPD to an A
600
of 1.0 were harvested, washed and

pretreated with 2-mercaptoethanol [43]. For spheroblasting,
cells were resuspended in 2 mL S-buffer (1.0 m sorbitol,
25 mm KH
2
PO
4
pH 6.5, 0.4 mm CaCl
2
), Zymolyase
100.000 was added (25 UÆmL
)1
), and cells were incubated
with gentle shaking at 30 ° C for 1 h. Spheroblasts were cen-
trifuged at 3600 g for 10 min and washed once with S-buf-
fer. An aliquot was mixed with sample buffer for direct
lysis (total input). Spheroblasts were lysed by adding 5 vol-
umes of N-buffer (18% Ficoll, 20 mm KH
2
PO
4
pH 6.5,
0.5 mm CaCl
2
,1mm phenylmethanesulfonyl fluoride) and
vigorous vortex-mixing. After centrifugation at 2.500 g, the
supernatant was re-centrifuged at 20 000 g for 30 min. The
supernatant representing the cytoplasmic fraction was care-
fully removed; aliquots of the cytoplasmic fraction and the
pellet with the nuclear fraction were mixed with sample
buffer and separated by SDS ⁄ PAGE (7% gel).

To check if the mutant War1p variants are post-transla-
tionally modified upon stress, cultures were grown to an
A
600
of  1, and split in half; one half was treated with
8mm potassium sorbate for 30 min the other remained
untreated. Cell lysates from cultures with an A
600
of 0.5
and 1.5 for War1-42p, were separated by SDS ⁄ PAGE (7%
gel), followed by immunoblotting using the monoclonal
HA antibody, 12CA5, to visualize War1-42p. Polyclonal
War1p antibodies were used to detect War1p-K762N,
War1p-F763M and War1p-R764D. Cycloheximide-chase
experiments were performed to analyze changes in protein
stability in the War1p mutants. Cultures were grown to an
A
600
of  1, and cycloheximide was added at a final concen-
tration of 0.1 mgÆmL
)1
. Equal amounts of cells were
harvested at the indicated time points, and cell extracts ana-
lyzed by immunoblotting as described above.
In vivo footprinting experiments and DNA
sequencing
In vivo footprints were performed exactly as described pre-
viously [44]. Cells were grown in 500 mL YPD to the early
exponential growth phase, concentrated in 10 mL YPD,
and treated with 5 lL dimethyl sulfoxide. The reaction was

stopped after 5 min, and chromosomal DNA prepared. Pri-
mer extension was carried out with a
32
P-labeled oligonu-
cleotide corresponding to the residues )497 to )472 of the
PDR12 promoter, resolved through a 8% sequencing gel,
exposed to a phosphoimager screen, and quantified. Traces
were captured using ImageQuant software, converted into
vector graphs, and aligned. For identification of the muta-
tions in war1 alleles, the WAR1 gene was amplified from
genomic DNA. To eliminate the danger of PCR-derived
mutations, the pool of PCR fragments was subjected
directly to DNA sequencing using the BigDye Terminator
Cycle Sequencing Kit version 3.0 according to the instruc-
tions of the manufacturer and the ABI PRISM Sequencing
System 310 (Applied Biosystems, Foster City, CA).
Acknowledgements
We thank Manuela Schu
¨
tzer-Mu
¨
hlbauer and all labor-
atory members for critical reading of the manuscript
and helpful discussions. Peter Piper and Mehdi Molla-
pour are acknowledged for sharing unpublished infor-
mation and their long-standing collaboration. We
appreciate the gifts of polyclonal anti-Swi6p and anti-
Hxk1p sera from Kim Nasmyth and Rudolf Schweyen,
respectively. This work was supported by a grant
from the FWF (Austrian Science Foundation Project

P-15934-B12) to K.K.
References
1 Lambert RJ & Stratford M (1999) Weak-acid preserva-
tives: modelling microbial inhibition and response.
J Appl Microbiol 86, 157–164.
2 Chichester DF & Tanner FW (1972) Antimicrobial food
additives. In Handbook of Food Additives (Furia TE,
ed.), pp. 115–184. CRC Press, Cleveland, OH.
C. Gregori et al. Yeast weak organic acid stress adaptation
FEBS Journal 274 (2007) 3094–3107 ª 2007 The Authors Journal compilation ª 2007 FEBS 3105
3 Krebs HA, Wiggins D, Stubbs M, Sols A & Bedoya F
(1983) Studies on the mechanism of the antifungal
action of benzoate. Biochem J 214, 657–663.
4 Stratford M & Anslow PA (1998) Evidence that sorbic
acid does not inhibit yeast as a classic ‘weak acid preser-
vative’. Lett Appl Microbiol 27, 203–206.
5 Piper PW (1999) Yeast superoxide dismutase mutants
reveal a pro-oxidant action of weak organic acid food
preservatives. Free Radic Biol Med 27, 1219–1227.
6 Piper P, Calderon CO, Hatzixanthis K & Mollapour M
(2001) Weak acid adaptation: the stress response that
confers yeasts with resistance to organic acid food pre-
servatives. Microbiology 147, 2635–2642.
7 Piper P, Mahe
´
Y, Thompson S, Pandjaitan R, Holyoak
C, Egner R, Mu
¨
hlbauer M, Coote P & Kuchler K
(1998) The Pdr12 ABC transporter is required for the

development of weak organic acid resistance in yeast.
EMBO J 17, 4257–4265.
8 Holyoak CD, Stratford M, McMullin Z, Cole MB,
Crimmins K, Brown AJ & Coote PJ (1996) Activity of
the plasma membrane H
(+)
-ATPase and optimal glyco-
lytic flux are required for rapid adaptation and growth
of Saccharomyces cerevisiae in the presence of the weak
acid preservative sorbic acid. Appl Environ Microbiol 62,
3158–3164.
9 Piper PW, Ortiz-Calderon C, Holyoak C, Coote P &
Cole M (1997) Hsp30, the integral plasma membrane
heat shock protein of Saccharomyces cerevisiae,isa
stress-inducible regulator of plasma membrane H
(+)
-
ATPase. Cell Stress Chaperones 2, 12–24.
10 Wolfger H, Mamnun YM & Kuchler K (2001) Fungal
ABC proteins: pleiotropic drug resistance, stress
response and cellular detoxification. Res Microbiol 152,
375–389.
11 Schu
¨
ller C, Bauer BE & Kuchler K (2003) Inventory
and evolution of fungal ABC protein genes. In ABC
Proteins from Bacteria to Man (Holland SM, Coole S,
Kuchler K & Higgins C, eds), pp. 279–293. Academic
Press, Elsevier Science, London.
12 Hazelwood LA, Tai SL, Boer VM, de Winde JH, Pronk

JT & Daran JM (2006) A new physiological role for
Pdr12p in Saccharomyces cerevisiae: export of aromatic
and branched-chain organic acids produced in amino
acid catabolism. FEMS Yeast Res 6, 937–945.
13 Papadimitriou MN, Resende C, Kuchler K & Brul S
(2007) High Pdr12 levels in spoilage yeast (Saccharo-
myces cerevisiae) correlate directly with sorbic acid
levels in the culture medium but are not sufficient to
provide cells with acquired resistance to the food preser-
vative. Int J Food Microbiol 113, 173–179.
14 Winzeler EA, Shoemaker DD, Astromoff A, Liang H,
Anderson K, Andre B, Bangham R, Benito R, Boeke
JD, Bussey H, et al. (1999) Functional characterization
of the S. cerevisiae genome by gene deletion and parallel
analysis. Science 285, 901–906.
15 Kren A, Mamnun YM, Bauer BE, Schu
¨
ller C, Wolfger
H, Hatzixanthis K, Mollapour M, Gregori C, Piper P &
Kuchler K (2003) War1p, a novel transcription factor
controlling weak acid stress response in yeast. Mol Cell
Biol 23, 1775–1785.
16 Todd RB & Andrianopoulos A (1997) Evolution of a
fungal regulatory gene family: the Zn (II) 2Cys6 binuc-
lear cluster DNA binding motif. Fungal Genet Biol 21,
388–405.
17 Friden P & Schimmel P (1987) LEU3 of Saccharomyces
cerevisiae encodes a factor for control of RNA levels
of a group of leucine-specific genes. Mol Cell Biol 7,
2708–2717.

18 Lohr D, Venkov P & Zlatanova J (1995) Transcrip-
tional regulation in the yeast GAL gene family: a com-
plex genetic network. FASEB J 9, 777–787.
19 Rottensteiner H, Kal AJ, Hamilton B, Ruis H & Tabak
HF (1997) A heterodimer of the Zn
2
Cys
6
transcription
factors Pip2p and Oaf1p controls induction of genes
encoding peroxisomal proteins in Saccharomyces cerevi-
siae. Eur J Biochem 247, 776–783.
20 Karpichev IV & Small GM (1998) Global regulatory
functions of Oaf1p and Pip2p (Oaf2p), transcription
factors that regulate genes encoding peroxisomal pro-
teins in Saccharomyces cerevisia. Mol Cell Biol 18,
6560–6570.
21 Creusot F, Verdiere J, Gaisne M & Slonimski PP (1988)
CYP1 (HAP1) regulator of oxygen-dependent gene
expression in yeast. I. Overall organization of the pro-
tein sequence displays several novel structural domains.
J Mol Biol 204, 263–276.
22 Pfeifer K, Kim KS, Kogan S & Guarente L (1989)
Functional dissection and sequence of yeast HAP1 acti-
vator. Cell 56, 291–301.
23 Bissinger PH & Kuchler K (1994) Molecular cloning
and expression of the Saccharomyces cerevisiae
STS1 gene product. A yeast ABC transporter
conferring mycotoxin resistance. J Biol Chem 269,
4180–4186.

24 Balzi E, Wang M, Leterme S, Van Dyck L &
Goffeau A (1994) PDR5, a novel yeast multidrug
resistance conferring transporter controlled by the
transcription regulator PDR1. J Biol Chem 269,
2206–2214.
25 Servos J, Haase E & Brendel M (1993) Gene SNQ2 of
Saccharomyces cerevisiae, which confers resistance to
4- nitroquinoline-N-oxide and other chemicals, encodes
a 169 kDa protein homologous to ATP-dependent per-
meases. Mol Gen Genet 236, 214–218.
26 Mahe
´
Y, Parle-McDermott A, Nourani A, Delahodde
A, Lamprecht A & Kuchler K (1996) The ATP-
binding cassette multidrug transporter Snq2 of
Saccharomyces cerevisiae: a novel target for the trans-
cription factors Pdr1 and Pdr3. Mol Microbiol 20,
109–117.
Yeast weak organic acid stress adaptation C. Gregori et al.
3106 FEBS Journal 274 (2007) 3094–3107 ª 2007 The Authors Journal compilation ª 2007 FEBS
27 Katzmann DJ, Hallstro
¨
m TC, Voet M, Wysock W,
Golin J, Volckaert G & Moye-Rowley WS (1995)
Expression of an ATP-binding cassette transporter-
encoding gene (YOR1) is required for oligomycin
resistance in Saccharomyces cerevisiae. Mol Cell Biol 15,
6875–6883.
28 Akache B, Wu K & Turcotte B (2001) Phenotypic ana-
lysis of genes encoding yeast zinc cluster proteins.

Nucleic Acids Res 29, 2181–2190.
29 Friden P, Reynolds C & Schimmel P (1989) A large
internal deletion converts yeast LEU3 to a cons-
titutive transcriptional activator. Mol Cell Biol 9,
4056–4060.
30 Nourani A, Papajova D, Delahodde A, Jacq C & Subik
J (1997) Clustered amino acid substitutions in the yeast
transcription regulator Pdr3p increase pleiotropic drug
resistance and identify a new central regulatory domain.
Mol Gen Genet 256, 397–405.
31 Lebel K, MacPherson S & Turcotte B (2006) New
tools for phenotypic analysis in Candida albicans: the
WAR1 gene confers resistance to sorbate. Yeast 23,
249–259.
32 Schu
¨
ller C, Mamnun YM, Mollapour M, Krapf G,
Schuster M, Bauer BE, Piper PW & Kuchler K (2004)
Global phenotypic analysis and transcriptional profiling
defines the weak acid stress response regulon in Sac-
charomyces cerevisiae. Mol Biol Cell 15, 706–720.
33 Schjerling P & Holmberg S (1996) Comparative amino
acid sequence analysis of the C6 zinc cluster family
of transcriptional regulators. Nucleic Acids Res 24,
4599–4607.
34 Vashee S, Xu H, Johnston SA & Kodadek T (1993)
How do ‘Zn2Cys6’ proteins distinguish between similar
upstream activation sites? Comparison of the DNA-
binding specificity of the GAL4 protein in vitro and
in vivo. J Biol Chem 268, 24699–24706.

35 Kaiser C, Michaelis S & Mitchell A (1994) Methods in
yeast genetics. A Laboratory Course Manual. Cold
Spring Harbor Laboratory Press, Cold Spring Harbor,
NY.
36 Wach A, Brachat A, Alberti-Segui C, Rebischung C &
Philippsen P (1997) Heterologous HIS3 marker and
GFP reporter modules for PCR-targeting in Saccharo-
myces cerevisiae. Yeast 13, 1065–1075.
37 Longtine, MS, McKenzie A, 3rd Demarini DJ, Shah
NG, Wach A, Brachat A, Philippsen P & Pringle JR
(1998) Additional modules for versatile and economical
PCR-based gene deletion and modification in Saccharo-
myces cerevisiae. Yeast 14, 953–961.
38 Myers AM, Tzagoloff A, Kinney DM & Lusty CJ
(1986) Yeast shuttle and integrative vectors with multi-
ple cloning sites suitable for construction of lacZ
fusions. Gene 45, 299–310.
39 Sikorski RS & Hieter P (1989) A system of shuttle vec-
tors and yeast host strains designed for efficient manipu-
lation of DNA in Saccharomyces cerevisiae. Genetics
122, 19–27.
40 Sathe GM, O’Brien S, McLaughlin MM, Watson F &
Livi GP (1991) Use of polymerase chain reaction for
rapid detection of gene insertions in whole yeast cells.
Nucleic Acids Res 19, 4775.
41 Egner R & Kuchler K (1996) The yeast multidrug trans-
porter Pdr5 of the plasma membrane is ubiquitinated
prior to endocytosis and degradation in the vacuole.
FEBS Lett 378, 177–181.
42 Mamnun YM, Pandjaitan R, Mahe

´
Y, Delahodde A &
Kuchler K (2002) The yeast zinc finger regulators Pdr1p
and Pdr3p control pleiotropic drug resistance (PDR) as
homo- and heterodimers in vivo. Mol Microbiol 46,
1429–1440.
43 Lohr D (1988) Isolation of yeast nuclei and chromatin
for studies of transcription-related processes. In Yeast, a
Practical Approach. IRL Press, Oxford.
44 Go
¨
rner W, Durchschlag E, Martinez-Pastor MT,
Estruch F, Ammerer G, Hamilton B, Ruis H & Schu
¨
ller
C (1998) Nuclear localization of the C
2
H
2
zinc finger
protein Msn2p is regulated by stress and protein kinase
A activity. Genes Dev 12, 586–597.
45 Rothstein RJ (1983) One-step gene disruption in yeast.
Methods Enzymol 101, 202–211.
C. Gregori et al. Yeast weak organic acid stress adaptation
FEBS Journal 274 (2007) 3094–3107 ª 2007 The Authors Journal compilation ª 2007 FEBS 3107

×