ISC1-encoded inositol phosphosphingolipid phospholipase C
is involved in Na
+
/Li
+
halotolerance of
Saccharomyces cerevisiae
Christian Betz
1
, Dirk Zajonc
1
, Matthias Moll
2
and Eckhart Schweizer
1
1
Lehrstuhl fu
¨r Biochemie and the
2
Lehrstuhl fu
¨r Anorganische und Allgemeine Chemie, Universita
¨t Erlangen-Nu
¨rnberg,
Erlangen, Germany
In Saccharomyces cerevisiae, toxic concentrations of Na
+
or Li
+
ions induce the expression of the cation-extrusion
ATPase gene, ENA1. Several well-studied signal transduc-
tion pathways are known correlating high salinity to the
transcriptional activation of ENA1. Nevertheless, informa-
tion on the actual sensing mechanism initiating these path-
ways is limited. Here, we report that the ISC1-encoded
phosphosphingolipid-specific phospholipase C appears to be
involved in stimulation of ENA1 expression and, conse-
quently, in mediating Na
+
and Li
+
tolerance in yeast.
Deletion of ISC1 distinctly decreased cellular Na
+
and Li
+
tolerance as growth of the Disc1::HIS5 mutant, DZY1, was
severely impaired by 0.5
M
NaCl or 0.01
M
LiCl. In con-
trast, K
+
tolerance and general osmostress regulation
were unaffected. Isc1Dmutant growth with 0.9
M
KCl and
glycerol accumulation in the presence of 0.9
M
NaCl or
1.5
M
sorbitol were comparable to that of the wild-type.
ENA1-lacZ reporter studies suggested that the increased salt
sensitivity of the isc1Dmutant is related to a significant
reduction of Na
+
/Li
+
-stimulated ENA1 expression. Cor-
respondingly, Ena1p-dependent extrusion of Na
+
/Li
+
ions
was less efficient in the isc1Dmutant than in wild-type cells.
It is suggested that ISC1-dependent hydrolysis of an
unidentified yeast inositol phosphosphingolipid represents
an early event in one of the salt-induced signalling pathways
of ENA1 transcriptional activation.
Keywords: salt-stress; signaling; sphingolipids; sphingolipid
phospholipase C; yeast.
The Saccharomyces cerevisiae gene, ISC1, has recently been
shown to encode an inositol phosphosphingolipid-specific
phospholipase C [1]. In vitro, the enzyme exhibits the
characteristics of a Mg
2+
-dependent neutral (N) sphing-
omyelinase (SMase) and, thus, resembles the most prom-
inent member of the SMase family present in mammalian
cells [2,3]. According to current knowledge, sphingomyelin
is absent from yeast and, hence, the physiological substrate
of Isc1p is likely to belong to one of the three major classes
of yeast sphingolipids, i.e. inositol phosphorylceramides,
mannositol phosphorylceramides, or mannosyldiinositol
phosphorylceramides [4]. In mammalian systems, various
intermediates of sphingolipid metabolism act as mediators
of intracellular signalling pathways [5–8]. In particular, the
SMase reaction product, ceramide, has been recognized as
a second messenger being induced by a variety of extracel-
lular stress signals [8,9]. Subsequent interaction of ceramide
with specific protein kinases, protein phosphatases or
proteinases induces signalling cascades which finally affect
basic cellular functions such as cell cycle progression, cell
growth, differentiation, apoptosis or Ca
2+
ion homeostasis
[8,9]. In S. cerevisiae, sphingolipids represent 20–30% of
cellular phospholipids [4] and, thereby, obviously fulfil an
important structural function. Besides this, they probably
contribute to the signal transduction potential of yeast cells,
too [10–15]. Their vital function is underlined by the
lethality of yeast mutants defective in sphingosine base
biosynthesis [16]. Although sphingosine base-defective
mutants may be partly suppressed by the production of
C26-fatty acid-containing glycerolipids, these mutants
remain sensitive against heat, osmotic and low pH stresses
[4,5,17]. From these results, the involvement of sphingoli-
pids in distinct stress response pathways of yeast became
quite obvious. Each one of various different stress
responses appears to have its own specific signalling
pathway [5]. While heat shock induces the biosynthesis
of trehalose [18,19], high extracellular osmolarity either
induces the accumulation of glycerol as a compatible
intracellular osmolyte [20–22] or, with toxic concentrations
of Na
+
or Li
+
ions, extrusion of these cations by induction
and activation of the specific, ATP-driven ion pump Ena1p
is initiated [21,23–26]. Both pathways of yeast osmoadap-
tation have been intensively studied and many of their
details are known. Non-specific osmostress is exerted by
moderate concentrations of various solutes such as NaCl,
KCl or sorbitol and induces the high-osmolarity glycerol
(HOG) pathway which rapidly raises the intracellular
glycerol concentration up to molar levels [20,21]. The
Correspondence to E. Schweizer, Lehrstuhl fu
¨r Biochemie;
Universita
¨t Erlangen, Staudtstrasse 5, D-91058 Erlangen, Germany.
Fax: +49 9131 8528254, Tel.: +49 9131 8528255,
E-mail: eschweiz@biologie.uni-erlangen.de
Abbreviations: (N-)SMase, (neutral)sphingomyelinase; HOG, high-
osmolarity glycerol; BSM, BODIPYFL-C
5
N-(4,4-difluoro-5,7-
dimethyl-4-bora-3a,4a-diaza-s-indacene-3-pentanoyl) sphingomyelin;
B-ceramide, BODIPYFL C
5
-ceramide; YPD, yeast extract,
peptone, dextrose; SCD, synthetic complete, dextrose.
Proteins and enzymes:Ena1 (atn1_yeast; EC 3.6.3.7), Isc1 (isc1_yeast;
EC 3.1.4.-), Gpd1 (g3p1_yeast; EC 1.2.1.12), Gpp2 (gpp2_yeast;
EC 3.1.3.-).
(Received 8 May 2002, revised 26 June 2002, accepted 5 July 2002)
Eur. J. Biochem. 269, 4033–4039 (2002) FEBS 2002 doi:10.1046/j.1432-1033.2002.03096.x
pathway comprises a mitogen-activated protein kinase
cascade which finally initiates transcription of the glycer-
ol-3-phosphate dehydrogenase (GPD1) and glycerol-3-
phosphatase (GPP2) genes [20]. In contrast, specific
halotolerance of yeast against extracellular NaCl or LiCl
is based on the induction of the ENA1/PMR2 gene
(designated as ENA1 in the following) encoding an enzyme
of the P-type ATPase family. This cation-extrusion pump
promotes the efflux of Na
+
and Li
+
from the cell. ENA1
expression is controlled by various different signalling
pathways [20–29]. Salt-stress-dependent induction of ENA1
involves the Ca
2+
/calmodulin-activated protein phospha-
tase calcineurin [25], the TOR-GLN3 signalling pathway
[27] and possibly also an additional, calcineurin-indepen-
dent mechanism [24]. Besides this, the alkaline response
regulator Rim101p [28] as well as glucose repression and
the HOG pathway contribute to ENA1 expression [20–
22,26,29]. While many details, mostly of the downstream
parts of these pathways, have been elucidated, little is
known about the sensing mechanisms and the signalling
molecules involved. Since in mammalian systems,
N-SMase has been recognized as a prominent effector of
sphingolipid-dependent stress responses [8,9,30], we were
interested to study whether, in yeast, the N-SMase
homologue, Isc1p, possibly serves as a stress signalling
mediator too. Here, we report that ISC1 is required for the
development of yeast halotolerance against Na
+
and Li
+
ions by means of HOG-independent induction of ENA1
expression.
EXPERIMENTAL PROCEDURES
Yeast strains, plasmids, chemicals and media
The yeast strains used in this study were JS91-15.23 (Mata,
ura3,trp1,his3,can1)andtheISC1 deletion strain DZY1
derived from it (MATa,Disc1::HIS5,ura3,trp1,his3,can1).
The HIS5 insertion cassette used for ISC1 disruption was
isolated, by short flanking homology PCR, from plasmid
pFA6a-HIS3MX containing the Schizosaccharomyces
pombe HIS5 gene [31]. The cassette exhibits, at both ends,
40 nucleotides of homology to positions 477–517 and 911–
951 of the ISC1 gene, respectively. The ENA1 ORF was
isolated by PCR amplification of two adjacent regions of
S. cerevisiae chromosomal DNA representing base pairs 1–
1182 and 1129–3273 of the ENA1 DNA sequence. The two
fragments were ligated by means of two overlapping,
terminal BamHI sites and subsequently inserted, as a PvuII/
XhoI fragment, between the ADH1 promoter and termina-
tor regions of the multicopy yeast expression vector,
pVT100-U [32]. The resulting plasmid was pCWB20.
Plasmid pDZ6 contained the ISC1 reading frame fused to
the MET25 promoter in the multicopy yeast expression
vector, p425MET25 [33]. Plasmid pFR70 containing an
ENA1-lacZ promoter–reporter fusion was obtained from
Prof. Rodriguez-Navarro, Madrid, Spain. Bacillus cereus
sphingomyelinase (SMase) was purchased from Sigma. The
fluorescent probes, BODIPYFL C5-sphingomyelin
(BSM) and BODIPYFL C5-ceramide (B-ceramide) were
from Molecular Probes Inc. Complex (YPD) and synthetic
complete (SCD) yeast media as well as the appropriate SCD
omission media were prepared according to standard
protocols [34].
Sphingomyelinase assay
The assay followed essentially the procedure described by
Ella et al. [35]. Yeast cells suspended in 1 vol. lysis buffer
(20 m
M
Tris/HCl pH 7.4, 10% glycerol, 50 m
M
KCl, 1 m
M
dithiothreitol, 1 m
M
phenylmethanesulfonyl fluoride, 1 l
M
pepstatin A, 10 l
M
leupeptin) were disrupted with glass
beads. Unbroken cells were removed by 5 min centrifuga-
tion at 4000 g. Total membranes were collected from the
supernatant by centrifugation at 100 000 gfor 1 h. The
fluorescent substrate, BSM, was subsequently used in a
semiquantitative SMase assay. Briefly, 55 lL of the mem-
brane suspension were mixed with 45 lL10m
M
Mes/KOH
buffer pH 6.0 containing 400 m
M
KCl, 200 l
M
BSM,
10 m
M
MgCl
2
,30m
M
2-glycerophosphate. After 1 h incu-
bation at 30 C, reaction products were extracted with
chloroform/methanol (1 : 1) and separated by TLC on
Silica G60 plates (Merck). The plates were developed with
chloroform/methanol/water (60 : 35 : 8) and, subsequently,
fluorescent spots were visualized and documented by a
Fluorescence Binocular (Zeiss, Stemi SV11, 515–565 nm
filter) according to the manufacturer’s indications.
Glycerol determination
Intracellular glycerol was determined enzymatically [36]
using a commercial glycerol determination kit (Roche
Diagnostics). Briefly, cells from mid-logarithmic phase
cultures of wild-type or isc1Dstrains were transferred, at
an D
600
of 0.8, from normal YPD media to YPD containing
0.9
M
NaCl. After 2 h growth at 30 C, cells were harvested
by centrifugation, washed twice with isotonic saline at 4 C
and then placed into boiling 0.5
M
Tris/HCl pH 7.0 for
10 min. After removing cell debris by 10 min centrifugation
at 15000 g, the glycerol in the supernatant was determined
enzymatically.
Measurement of intracellular Na
+
and Li
+
concentrations
Determinations followed essentially the procedure described
by Gaxiola et al. [37]. In brief, cells were grown in YPD
media to the densities indicated for the particular experi-
ment. After harvesting by centrifugation, cells were washed
three times with 1.5
M
sorbitol. For subsequent cell
extraction, two alternative methods were used. Method A:
cells were permeabilized by incubation for 15 min at 95 C.
Na
+
and Li
+
were determined in the cleared extracts using
Na
+
and Li
+
specific lamps (L.O.T.-Oriel GmbH, Darms-
tadt, Germany) in a Shimadzu AA-6200 atomic absorption
flame emission spectrophotometer. Method B: cells were
washed and lyophilized. The dry cells were incinerated at
840 C for 6 h. The residue was dissolved in 0.1
N
HCl and
atomic absorption measurements were performed as
described under method A.
RESULTS
S. cerevisiae ISC1
mutants are sensitive to Na
+
and Li
+
ion stresses
According to Sawai et al. [1] disruption of the yeast ORF,
ISC1, abolishes the in vitro SMase activity of the wild-type
4034 C. Betz et al. (Eur. J. Biochem. 269)FEBS 2002
cell homogenate. The characteristics of the Disc1::HIS5
deletion strain, DZY1, which was constructed in this work
are in accordance with these findings (Fig. 1). SMase
activity was efficiently restored in isc1Dcells upon transfor-
mation with plasmid pDZ6 encoding the intact ISC1 gene
(Fig. 1). Comparable growth rates were observed with wild-
type and isc1Dcells in normal YPD media not only at 30 C
but also at elevated temperature (37 C) or low pH (pH 3.5)
stresses (Fig. 2A). However, in the presence of 0.4–0.9
M
NaCl, growth of the mutant was differentially reduced
(Fig. 2B) and wild-type cells rapidly overgrew the mutants
(Fig. 2A). After eight generations in 0.9
M
NaCl, the
proportion of isc1Dcells had dropped to 2% of the
viable cells, which compares to > 80% isc1Dcells surviving
in the absence of NaCl under otherwise identical conditions
(Fig. 2A). On solid media, the differential sensitivity of
isc1Dcells against elevated (0.4–0.5
M
) NaCl concentration,
was further confirmed and, in addition, a similar toxicity
was established for 0.01
M
LiCl (Fig. 3). In contrast, 0.8
M
KCl had no measurable inhibitory effect on isc1Dgrowth
on solid media (Fig. 3).
ISC1
functions independently of the HOG-pathway
Adaptation of yeast to high salinity is, according to current
knowledge, largely based on two different mechanisms, i.e.
induction of the HOG pathway responding to nonspecific
osmostress [20–23], and induction of the ion extrusion pump
Ena1p responding to toxic concentrations of Na
+
and Li
+
ions [21,25–29]. According to the data shown in Figs 2 and
3, isc1Dcells are specifically sensitive to NaCl and LiCl, but
tolerate high osmolarity of other solutes such as KCl
(Fig. 3) or glucose (data not shown). These characteristics
argue against the HOG pathway being affected in the isc1D
mutant. In agreement with this conclusion, cellular glycerol
levels increased to comparable levels in wild-type and isc1D
cells upon raising the salinity and osmolarity of the media
(Table 1). Thus, the HOG signalling pathway responded
normally in the mutant not only with 1.5
M
sorbitol but also
with 0.9
M
NaCl.
ISC1
is involved in Na
+
and Li
+
salt-induced expression
of
ENA1
Stimulation of ENA1 expression has been recognized as a
crucial response of yeast to extracellular high salinity [20–
29]. The ENA1 encoded ATPase mediates Na
+
and Li
+
ion
extrusion from the cell. We therefore investigated whether
the loss of halotolerance in isc1Dcells was due to the failure
of ENA1 induction in the mutant. For this, the ENA1-lacZ
promoter–reporter construct in plasmid pFR70 was trans-
formed into wild-type and isc1Dcells. The transformants
expressing the bacterial lacZ gene under the control of the
ENA1 promoter were challenged with 0.8
M
KCl, 0.8
M
NaCl and 0.25
M
LiCl, respectively. In the wild-type
transformants, increasing concentrations of NaCl and LiCl
caused the expected time- and concentration-dependent,
strong induction of b-galactosidase activity (Fig. 4). In the
isc1Dtransformants, however, b-galactosidase induction
Fig. 1. SMase activity in wild-type (JS91-15.23) and ISC1-disrupted
yeast cells. From each strain 550 lgmembraneproteinwereapplied
to the fluorescent SMase assay as described in Experimental proce-
dures. Purified Bacillus cereus SMase (0.1 U) was used in a control
assay. The fluorescent sphingomyelin derivative BSM and its SMase
product, B-ceramide were run as references. isc1D+pDZ6 was a
transformant of the isc1Dmutant with the ISC1 containing plasmid,
pDZ6.
Fig. 2. Differential growth rates of wild-type
and isc1Dcells under different stress conditions.
(A) One-to-one mixtures of wild-type (JS91-
15.23) and isc1Dcells were inoculated into
SCD media and subsequently incubated under
the following conditions: 30 C(s), 37 C
(.), pH 3.5 (h), with 0.9
M
NaCl (d). Both
strains had been precultivated in SCD media
up to mid-log phase. At distinct time intervals,
aliquots of each culture were withdrawn and
plated onto SCD media. After outgrowth the
cells were replica-plated onto histidine-omis-
sion media. The ratio of histidine-positive
isc1Dcells to nondisrupted, histidine-requiring
JS91-15.23 cells was then determined for each
sample. (B) JS91-15.23 (d)andisc1D(s)cells
were grown separately in YPD media con-
taining 0.4
M
NaCl. Identical cell counts were
used for inoculation of the two strains.
FEBS 2002 Salt-stress signalling in yeast (Eur. J. Biochem. 269) 4035