Curr Genet (2001) 39: 127±136

DOI 10.1007/s002940100189

O R I GI N A L P A P E R

P. Ververidis á F. Davrazou á G. Diallinas

D. Georgakopoulos á A. K. Kanellis á N. Panopoulos

A novel putative reductase (Cpd1p) and the multidrug exporter

Snq2p are involved in resistance to cercosporin and other singlet
oxygen-generating photosensitizers in Saccharomyces cerevisiae
Received: 5 September 2000 / Accepted: 18 December 2000 / Published online: 30 March 2001
Ó Springer-Verlag 2001

Abstract Phytopathogenic Cercospora species produce dependent pyridine nucleotide reductases. We showed
cercosporin, a photoactivated perylenequinone toxin that over-expression of either of these proteins can also
that belongs to a family of photosensitizers which mediate resistance to other singlet oxygen-generating
absorb light energy and produce extremely cytotoxic, compounds. The involvement of Snq2p and Cpd1p in
reactive oxygen species. In this work, we used photosensitizer detoxi®cation reinforces previous
Saccharomyces cerevisiae as a model system for the observations which suggested that singlet oxygen acts
identi®cation and cloning of genes whose products on membrane lipids and that cellular resistance to
mediate cercosporin detoxi®cation. Two genesexpressed cercosporin is mediated by a mechanism involving
in high-copy number vectors conferred cercosporin toxin e‚ux and/or toxin reduction.
resistance to an otherwise sensitive strain. One gene
codes for Snq2p, a well-characterized multidrug, ABC-
Keywords Cercosporin detoxi®cation á
type, e‚ux protein. The other, designated CPD1
Singlet oxygen-resistance á Gene overexpression á
(Cercosporin Photosensitizer Detoxi®cation), encodes a
Multidrug resistance
novel protein with signi®cant similarity to the FAD-

Introduction
Communicated by L. A. Grivell
Filamentous fungi of the genus Cercospora are patho-
P. Ververidis (&)
Department of Crop Science, Plant Biochemistry & genic to many economically important crops, among
Biotechnology Laboratory, School of Agricultural Technology, them banana, co€ee, corn, peanut, soybean, sugar beet
Technological Educational Institute, P.O. Box 140, and tobacco. Most phytopathogenic species of the
71110 Heraklion, Crete, Greece genus produce a lipid-soluble red polyketide (perylen-
E-mail: [email protected]
equinone) toxin known as cercosporin (Daub 1982,
P. Ververidis á F. Davrazou 1987). Work from several laboratories strongly sup-
D. Georgakopoulos á N. Panopoulos (&) ports a principal role for cercosporin in host plant
Institute of Molecular Biology and Biotechnology,
Foundation for Research and Technology Hellas, infection and virulence (Upchurch et al. 1991; Daub
P.O. Box 1527, 71110 Heraklion, Crete, Greece et al. 1992; Daub and Ehrenshaft 1995). Cercosporin is
E-mail: [email protected] a non-host-speci®c, photoactivated toxin which inter-
F. Davrazou á N. Panopoulos acts with molecular oxygen to produce reactive oxygen
Department of Biology, University of Crete, species (ROS), principally singlet oxygen (Spikes 1989)
71409 Heraklion, Crete, Greece or, under reducing conditions, radical forms of ROS.
G. Diallinas ROS toxicity has been implicated in a number of
Department of Biology, University of Athens, processes of plant and animal pathogenesis (Halliwell
Panepistimioupolis, 15781 Athens, Greece and Gutteridge 1990). Interestingly, ROS are also
A.K. Kanellis usedby animal and plant cells as defensive agents or
Department of Pharmaceutical Sciences, signals to induce defensive systems against pathogens
Aristotle University of Thessaloniki, 54006 Thessaloniki, Greece
(Mehdy et al. 1996). ROS, depending on the location
D. Georgakopoulos of their production, are able to react with all kinds of
Present address: School of Agricultural Technology, biomolecules (lipids, nucleic acids and proteins; Ito
Technological & Educational Institute,
P.O. Box 140, 71110 Heraklion, Crete, Greece 1981; Knox and Dodge 1985). However, when ROS
128

derive from an exogenous source, their primary target

site is the cell membrane, where they cause lipid Experimental methods
peroxidation, altered membrane function, electrolyte
Media and culture conditions
leakage and ultimately cell death (Daub 1982; Daub
and Briggs 1983; Knox and Dodge 1985; Daub and S. cerevisiae strains were grown either in rich (YPD) or minimal
Ehrenshaft 1995). (MYM) media with appropriate supplements when necessary, as
Cercosporin is toxic to a wide range of organisms described previously (Adams et al. 1997). Plasmid-borne drug
tested, including many bacteria, fungi, plants and ani- resistance was identi®ed by the 5-¯uoroorotic acid (5-FOA) test.
Eviction of the plasmid (containing one copy of the URA3 gene
mals (Daub 1987). However, a small number of bac- and one copy of the gene of interest) was monitored by screening
teria and fungi are relatively resistant to the toxin. The for colonies that were URA± and therefore resistant to the drug
highest degree of resistance observed is exhibited in 5-FOA. The 5-FOAr colonies were then screened for resistance
Cercospora species themselves, which can thrive in against cercosporin and were found to be sensitive (Ausubel et al.
1994). YPD and MYM media containing appropriate concentra-
cercosporin concentrations up to 1,000-fold higher than tions of photosensitizers [(Sigma): 1±20 lM cercosporin,
that which is lethal to most other organisms. The 25±100 lM methylene blue (MB), 25±100 lM toluidine blue (TB)]
mechanisms underlying cercosporin resistance by were prepared by addition of the drug, from 100-fold concentrated
Cercospora species or other microbes are not fully stock organic solvent solutions, in warm (45 °C) media. Appro-
priate organic solvent volume additions were added to the controls
understood. In most cases, cercosporin resistance is (no photosensitizer drug present). Yeast strains to be tested for
associated with general resistance to activated singlet drug resistance were grown to the exponential growth phase
oxygen species. These are also produced by other (OD600=0.5) and 5 ll of serial dilutions (1:1, 1:10, 1:100, 1:1000) of
photosensitizers such as chlorophyll, porphyrins, cell cultures were spotted onto media supplemented with photo-
psoralens, ribo¯avin and synthetic dyes such as acridine sensitizers and incubated for 48 h (YPD plates) under 49.3 lmol
m±2s±1 of white illumination (400±500 nm).
orange, methylene blue, rose bengal and toluidine blue
(Daub and Ehrenshaft 1995; Ehrenshaft et al. 1998).
Yeast and bacterial strains
The best documented mechanism of defense against
photosensitizers is the production of carotenoids, which Standard Escherichia coli strains such as, DH5a, XL-1-Blue and
are the most ecient quenchers of singlet oxygen HB101 served as plasmid hosts, using the E. coli transformation
species known in nature (Young 1991). However, car- protocol described by Sambrook et al. 1989. The S. cerevisiae
otenoids do not mediate cercosporin and singlet oxygen strains used are shown in Table 1.
resistance in Cercospora species (Ehrenshaft et al.
1995). Furthermore, enzymes known to be e€ective Plasmids
against radical and reduced oxygen species (e.g. cata- Plasmids pRS316, pRS314, YEep352 and YEplac112 were the
lase, peroxidases and superoxide dismutase) are not standard S. cerevisiae vectors used in this work (Sikorski and
e€ective against cercosporin and singlet oxygen (Wil- Hieter 1989; Ausubel et al. 1994). The plasmid pCR1 was isolated
liamson and Scandalios 1993). Daub and her colleagues from the high-copy vector YEp352 library carrying open reading
have proposed a model for cercosporin self-resistance frames (ORFs) YNR074c (SGDID S0005357) and YNR073c
(SGDID S0005356; see Fig. 1). Plasmids pYNR074 (subsequently
in which the toxin is transiently and reversibly reduced renamed pCPD1hc) and pYNR073 were made by subcloning
by Cercospora hyphae (Daub et al. 1992;Sollod et al. restriction fragments SmaI (2.79 kb) and PstI (2.15 kb) respec-
1992). This model was initially based on biochemical tively, from pCR1 to YEp352. pCPD1lc was the low copy plasmid
and ¯uorescence microscopy studies showing that living pRS316 carrying CPD1 cloned as a SmaI (2.79 kb) restriction
fragment isolated from pCR1. pSNQ2hc (pEH22; Haase et al.
hyphae are capable of reducing cercosporin (Leishman 1992) and pSNQ2lc were plasmids YEp24 and pRS316 respec-
and Daub 1992); and it gained additional genetic sup- tively, carrying SNQ2. Plasmid pSNQ2lc was made by cloning a
port by the observation that cercosporin-sensitive 7.5 kb XhoI-SalI restriction fragment into the SalI site of pRS316.
mutants of C. nicotianae are unable to reduce cerco-
sporin (Jenns et al. 1995). Gene disruptions
Our laboratory is interested in the mechanisms un-
derlying cercosporin resistance. To study these mecha- Gene disruption was performed following the one-step, gene-
nisms, we sought to develop an approach for cloning the replacement procedure (Adams et al. 1997), as follows. Subcloning
genes that might be directly involved in cercosporin of the SmaI fragment of pCR1 in a pUC19 plasmid facilitated the
replacement of CPD1 sequences with URA3. Cloning a ®lled-in
detoxi®cation, using an easily manipulated eukaryotic HindIII 1,166 bp restriction fragment (containing the URA3 gene)
model system, such as Saccharomyces cerevisiae. Our into the EcoRV restriction site of CPD1 replaced the EcoRV
approach identi®ed two alternative mechanisms leading fragment of CPD1 (see Fig. 1D). Dcpd1 strain was constructed by
to cercosporin and other photosensitizer resistance. One transforming a YPH98 strain with a HindIII restriction fragment of
plasmid pDcpd1 that included the YNR074c locus (CPD1) in which
involves a well known multidrug ABC exporter protein the central region had been replaced with the URA3 gene. URA3
and the other a novel putative ¯avoprotein member of transformants were selected and correct integration of the deletion
the pyridine nucleotide-disul®de oxidoreductasefamily. construct was con®rmed either by Southern blot analysis or by
Our results strengthen previous suggestions that cerco- polymerase chain reaction analysis of genomic DNA isolated from
selected transformants (Adams et al. 1997). The yeast genomic
sporin is reduced in other resistant fungi; and they are high-copy vector YEp352 library was kindly provided by G. Thi-
suggestive of a mechanism for singlet oxygen detoxi®- reos (Institute of Molecular Biology and Biotechnology, Herak-
cation in eukaryotic cells. lion) and was used to transform and functionally complement
 129

Table 1 Saccharomyces cerevisiae strains used in this work

Strain Genotype Source/Reference

YPH98 MATa, ura3-52, lys2-801, ade2-101, leu2-D1, trp1-D1 K Kuchler and R Egner (see also Sikorski
and Hieter 1989)
YPH98(Dcpd1) MATa, ura3-52, lys2-801, ade2-101, leu2-D1, trp1-D1, This work
(cpd1::URA3)
YPH98 (Dsnq2) MATa, ura3-52, lys2-801, ade2-101, leu2-D1, trp1-D1, K Kuchler and R Egner (see also Sikorski and Hieter
[snq2::Tn10 (URA3)] 1989; Servos et al. 1993)
YPH500 MATa, ura3-52, lys2-801, ade2-101, trp1-D63, As above (see also Sikorski and Hieter 1989)
his3-D200, leu2-D1
YKKB13 MATa, ura3-52, lys2-801, ade2-101, trp-1D63, As above (see also Bissinger and Kuchler 1994)
his3-D200, leu2-D1, (pdr5::TRP1)
R757 MATa, ura3-52, his4-15, lys9 RF Gaber (see also Gaber et al. 1990)
R2903 MATa, ura3-52, his4-15, lys9, (Dhol1-1) As above
YHR048wt MATa, ade2, ura3, his3-D200, lys2-801, leu2-3,112, As above
trp1-903, tyr1-501, Kmr
(D28) MATa, ade2, ura3, his3-D200, lys2±801, leu2-3,112, As above
trp1-903, tyr1-501, Kms, (Dyhr048)
YHL040wt MATa, ade2, ura3, his3-D200, lys2±801, leu2-3,112, As above
trp1-903, tyr1-501, Kmr
(D170) MATa, ade2, ura3, his3-D200, lys2-801, leu2-3,112, As above
trp1-903, tyr1-501, Kms, (Dyhl040)
YHL047wt MATa, ade2, ura3, his3-D200, lys2-801, leu2-3,112, As above
trp1-903, tyr1-501, Kmr
(D163) MATa, ade2, ura3, his3-D200, lys2-801, leu2-3,112, As above
trp1-903, tyr1-501, Kms, (Dyhl047)

strain YPH98. Cercosporin-resistant transformants were isolated addition of amino acid, purine or pyrimidine supple-
as described by Nitiss and Wang (1988). ments in growth media were found not to a€ect sensi-
tivity to cercosporin (results not shown). For all further
Nucleic acid isolation, manipulation and hybridization experiments shown in this work, we chose the well
characterized and widely used laboratory strain
Plasmid isolations and manipulations were as described by Sam-
brook et al. (1989). Plasmid rescue from yeast transformants was as YPH98.
described by Robzyk and Kassir (1992). Yeast genomic DNA and We followed two alternative approaches to identify
total RNA were isolated by techniques described in Ausubel et al. S. cerevisiae genes that, upon expression from high copy
(1994) and Adams et al. (1997), respectively. All DNA and RNA plasmids, conferred cercosporin resistance to strain
manipulations wereas described in Sambrook et al. (1989).
YPH98. The ®rst aimed at selecting clone(s) from a high-
copy vector gene library that conferred cercosporin
Results resistance upon transformation of YPH98. The second
was based on the observation that a putative cercosporin
Rationale for selecting a cercosporin-resistant export protein (Cfp, identi®ed in C. kikuchii) was similar
phenotype in S. cerevisiae to Sge1p and Atr1p of S. cerevisiae, both members of
the major facilitator superfamily-multidrug resistance
We were interested in establishing an easily manipu- (MFS-MDR) family (Callahan et al 1999). In this
lated microbial system as a model organism to study method, we tested whether representatives of subfamilies
the mechanisms underlying the cytotoxic activity of of S. cerevisiae MDR transporters conferred cercosporin
cercosporin. We tested Aspergillus nidulans, E. coli, resistance when expressed from high-copy plasmids.
Phytophthora parasitica, Pseudomonas syringae and
S. cerevisiae for growth on MYM and YPD media
containing 1±20 lM cercosporin under light and dark Identi®cation of CPD1, a novel gene responsible
conditions. Our analysis showed that Ph. parasitica and for cercosporin resistance
S. cerevisiae exhibited the highest, light-dependent
sensitivity to the toxin (results not shown). The Strain YPH98 (for full genotype, see Table 1) was
advanced state of yeast genetics and molecular biology, transformed with a S. cerevisiae genomic library con-
combined with the availability of its entire genome structed in vector YEp352 (gift of G. Thireos), using the
sequence in databases, made S. cerevisiae an ideal standard lithium chloride technique (Ausubel et al.
working system for our purposes. Various laboratory 1994). Transformants capable of growing on cercosporin
strains of S. cerevisiae were found to be sensitive to were selected in two steps (see Materials and methods).
cercosporin concentrations over 1 lM on both MYM Among 700,000 transformed cells (URA3+) plated, nine
and YPD media. The presence of standard amino acid gave rise to cercosporin-resistant colonies (results not
or nucleobase auxotrophies in di€erent strains or the shown). All nine colonies were puri®ed and the possi-
130

bility that cercosporin resistance might be linked to cloned in high-copy vector YEp352 (for details on their
plasmid-borne sequences was tested, both by the 5-FOA construction, see Materials and methods). The corre-
vector eviction test (Ausubel et al. 1994) and by plasmid sponding plasmids, named pYNR074 and pYNR073
rescue and re-introduction to YPH98 (Robzyk and (see Fig. 1), were used to transform YPH98. Puri®ed
Kassir 1992). Both tests showed that, in three colonies, transformants were analyzed for their ability to grow in
cercosporin resistance was plasmid-borne. Restriction the presence of cercosporin. Only pYNR074 conferred
analysis of all three rescued plasmids, using EcoRI and cercosporin resistance in YPH98, to the same level as
HindIII, showed an identical restriction pattern (results that conferred by the original clone pCR1 (see Fig. 2).
not shown), suggesting that all three plasmids carried the Thus, the gene corresponding to ORF YNR074c was
same genomic insert of approximately 4.5 kb. This solely responsible for cercosporin resistance when
clone, named pCR1, was further analyzed. present in high-copy-number. This novel gene was
By partial sequence and in silico analysis, we identi- designated CPD1 (cercosporin and photosensitizer det-
®ed the genomic region in pCR1. This region corre- oxi®cation) and plasmid pYNR074 was renamed
sponds to part of chromosome XIV and includes two pCPD1hc.
ORFs called YNR074c and YNR073c (Fig. 1). Both We also tested whether cercosporin resistance was
ORFs and their ¯anking sequences were subsequently dependent on CPD1 copy number. Strain YPH98 was
transformed with pCPD1lc, a low-copy plasmid carrying
CPD1 (see Materials and methods). This plasmid con-
ferred low level resistance to YPH98 (results not shown).

CPD1 is not essential for yeast growth

but is necessary for resistance to cercosporin

To further investigate the role of CPD1, the gene was

disrupted in S. cerevisiae strain YPH98. For this pur-
pose, plasmid pDcpd1 was constructed as described in
Materials and methods. YPH98 strain was transformed
with the HindIII fragment of pDcpd1 and transformants
with the correct integration of the deletion construct, as
judged by Southern blot analysis, were selected (results
not shown). cpd1::URA3 mutants showed normal
growth rates on both MYM and YPD media. In con-
trast, cpd1 null mutants showed increased sensitivity to
cercosporin (see Fig. 2).

In silico analysis reveals that CPD1 is a putative

transmembrane reductase

The CPD1 ORF encodes a predicted protein product of

378 amino acids (mol. wt. 41,335; Fig. 3). We identi®ed

Fig. 2 Cercosporin resistance pro®le of YPH98 yeast strains

overexpressing or not expressing CPD1. CPD1 is the original
YPH98 strain transformed with plasmid pRS316 (URA3+).
hc-CPD1 is a YPH98 strain transformed with pCPD1hc. Dcpd1 is
Fig. 1 Physical map of plasmid constructs: A pCR1, B pCPD1hc a YPH98 strain carrying a null allele of CPD1. Equal amounts of
(pYNR074), C pYNR073 clones and D pDcpd1 (for further details serial dilutions of cells from exponentially growing cultures
see Materials and methods). hc denotes high copy number, (+) and (OD600=0.5) were spotted onto plates of yeast extract/peptone/
(±) represent plasmids which confer or do not confer cercosporin dextrose (YPD) medium containing either no toxin, 2 lM, or 5 lM
resistance, respectively. Restriction site abbreviations: B BamHI, E cercosporin, and allowed to grow for 48 h at 30 °C under
EcoRI, EV EcoRV, H HindIII, P PstI, S SmaI 49.3 lmol m±2 s±1 of visible light (400±700 nm)
 131

Fig. 3 Nucleotide sequence of CPD1 present in plasmid pCR1 and the general family of FAD-dependent pyridine nucleo-
deduced amino acid sequence of the Cpd1p protein. Putative tide reductases (FADPNR). Most of these proteins are
membrane associated segments are shadowed, while nucleotide-
binding domains (FAD/NAD(P)H) are underlined predicted to have 1±3 strong hydrophobic segments,
suggestive of putative transmembrane domains.
Although the overall similarity of Cpd1p with these
several possible Cpd1p homologues from a number enzymes is not very high [26±27% identical amino acid
of other organisms in the databases. These proteins residues with trypanothione reductase (TR), 23±24%
appeared as ®rst strikes when searching the databases identity with glutathione reductase (GR), putidaredoxin
of di€erent organisms, using the whole Cpd1p peptide reductase (PRX) and rubredoxin reductase (RRX), and
sequence as an in silico probe. AllCpd1p homologues 20±22% identity with other reductases], there are
appear to be classi®ed as oxidoreductases, belonging to conserved sequence motifs such as FAD-binding and
132

NAD(P)H-binding domains (Fig. 4). Interestingly, family. Assuming that Cpd1p and its homologues di-
Cpd1p does not appear to possess the disul®de redox verged from an ancestral FAD/NAD(P)H reductase
central domain found in all reductases of the PNR (Kuriyan et al. 1991), and that disul®de reductase ac-
 133
b
Fig. 4 Multiple alignment of amino acid sequences (domains A, B,
C, D) shared by Cpd1p and other FAD/NAD(P)H dependent
oxidoreductases. TRX Thioredoxin reductase [Escherichia coli
P09625 (Kuriyan et al. 1991), Saccharomyces cerevisiae P29509,
Human Q16881], GR glutathione reductase [Arapidopsis thaliana
P48641, E. coli (Tartaglia et al. 1990), Pseudomonas cepacia
P48639, S. cerevisiae P41921, Drosophila melanogaster P91938,
Human P00390 (Kuriyan et al. 1991)], TR trypanothione reductase
(Trypanosoma brucei P39051), AHR alkyl hydroperoxide reductase
[E. coli (Tartaglia et al. 1990)], SQR sul®de quinone reductase [Rho-
dobacter capsulatus (Schutz et al. 1997)], RRX rubredoxin reductase Fig. 5 Cercosporin resistance pro®le of YPH98 yeast strains
[P. oleovorans (Eggink et al. 1990)], PRX putidaredoxin reductase overexpressing or not expressing SNQ2. SNQ2 is the original
[P. putida (Eggink et al. 1990)], CPD1 cercosporin photosensitizer YPH98 strain transformed with plasmid pRS316 (URA3+). hc-
detoxi®cation (S. cerevisiae, this work). Proteins whose three- SNQ2 is a YPH98 strain transformed with pSNQ2hc. Dsnq2 is a
dimensional structure has been solved are marked with *. Amino YPH98 strain carrying a null allele of SNQ2. Equal amounts of
acids in bold indicate isofunctional/identical residues. The x in the serial dilutions of cells from exponentially growing cultures
consensus indicates that any residue is acceptable, while ± reveals a (OD600=0.5) were spotted on YPD plates containing either no
gap in sequence alignment. Consensus sequences are shadowed. a toxin, or 2 lM, or 5 lM cercosporin, and allowed to grow for 48 h
and b denote a-helix and b-sheet, respectively at 30 °C under 49.3 lmol m±2.s±1 of visible light (400±700 nm)

tivity emerged later during divergent evolution, our cant cercosporin resistance. Expression from a low-copy
analysis strongly suggests that Cpd1p is a homologous plasmid still enhanced the ability of the recipient strain
member of the FADPNR family. Further biochemical to grow on cercosporin (results not shown); and snq2
analysis is required to determine the exact redox mech- null mutants showed increased sensitivity to cercosporin,
anism and the Cpd1p enzyme characteristics. compared to a wild-type isogenic strain.

High-copy expression of SNQ2, an ABC multidrug CPD1 and SNQ2 confer singlet oxygen resistance
transporter, confers cercosporin resistance
In order to examine whether Cpd1p and Snq2p are
Yeast possesses two recognized families of MDR trans- required for general resistance against photosensitizers
porters; the ABC-MDR (ATP-binding cassette) and the that produce singlet oxygen, we tested yeast strains
MFS-MDR (major facilitator superfamily). Members of either over-expressing or not expressing CPD1 or SNQ2
the ®rst are energized by ATP, while members of the for their ability to grow in the presence of two other
second are energized by the membrane proton-motive synthetic singlet oxygen-generating photosensitizers,
force (for a review see Andre 1995; Paulsen et al. 1996; MB and TB. Our results, shown in Fig. 6, demonstrate
Go€eau et al. 1997). The physiological roles of these that both proteins (when over-expressed) conferred
proteins remain obscure. To test whether MDR trans- enhanced resistance against the toxic e€ects of both
porters are involved in cercosporin detoxi®cation, we dyes. Expression of CPD1 from low-copy plasmid con-
selected and analyzed the ability of representative mem- ferred resistance, albeit lower than when expressed
bers of both the ABC- and MFS-MDR yeast families to fromhigh-copy plasmid, to both dyes. Finally, cpd1 and
confer cercosporin resistance. snq2 null mutants were not visibly more sensitive to
The well characterized proteins of S. cerevisiae either dye than a wild-type control strain (see Fig. 6).
Snq2p and Pdr5p (ABC-MDR; Servos, et al. 1993; We also tested whether photosensitizer resistance
Balzi et al. 1994; Bissinger and Kuchler 1994; Decot- conferred by CPD1 and SNQ2 is additive. We
tignies et al. 1994, 1995; Hirata et al. 1994; Sanglard co-transformed YPH98 with plasmids carrying CPD1 or
et al. 1995; Kolaczkowski et al. 1996; Mahe et al. 1996) SNQ2, either in high-copy or in low-copy vectors. All
and those of Sge1p and Atr1p (MFS-MDR; Gaber et al. possible combinations were tested. Transformants were
1990; Ehrenhofer-Murray et al. 1994, 1998; Andre puri®ed and tested for their ability to grow on YPD
1995) were tested for their ability to confer, upon plates supplemented with cercosporin, MB, or TB. Our
overexpression, enhanced growth on cercosporin. Over- results showed that double CPD1/SNQ2 transformants
expressing strains were compared to corresponding were not more resistant to cercosporin than single
isogenic wild-type alleles and null mutants (results not transformants over-expressing either CPD1 or SNQ2
shown). We also tested whether null mutations in the (data not shown).
genes HOL1, YHR048w, YHL040c and YHL047c
(encoding representatives of other subfamilies of MFS-
MDR) confer increased sensitivity to cercosporin Discussion
(results not shown).
Among all proteins tested, we found that only Snq2p We were interested in developing an easily manipulated
is related to cercosporin resistance (Fig. 5). SNQ2 over- microbial system which would enable us to identify and
expression from a high-copy plasmid conferred signi®- clone genes that could confer resistance to photosensi-
134

performed here are suggestive of its topology and

function. The primary amino acid sequence, and par-
ticularly the short sequence motifs, suggest that Cpd1p
might function as a FAD pyridine nucleotide reductase
(Fig. 4). When the secondary and tertiary structure of
Cpd1p is modeled in silico, using various programs
available in the public domain, it shows the presence of
typical bab nucleotide-binding motifs and other folding
patterns found in all four domains of other FAD-
dependent reductases with resolved structure, such as
GR, TR and THR (Wierenga et al. 1986). The function
of these reductases is suggestive of the possible physio-
logical function of Cpd1p. GR, found in most aerobic
organisms, catalyzes the reduction of oxidized glutathi-
one to two molecules of reduced glutathione (c-glu-
tamylcysteinylglycine), which is widely distributed in
microorganisms, plants and animals and has various
functions in defense against oxidative stress and xeno-
biotic toxicity (Meister and Anderson 1983; Izawa et al.
1998). TR, a parasitic homologue of GR, reduces an
oxidized bis-(glutathione) spermidine conjugate, trypa-
nothione [T(S)2; (Shames et al. 1986)]. THR reduces
thioredoxins, small redox-active proteins, which have
been shown to have diverse functions in bacteria
(Holmgren 1985). GR, TR and THR are all enzymes
important in the cell's defense against oxidative stress
(Meshnick et al. 1978; Arrick et al. 1981; Holmgren
1985; Muller 1991). This work strongly suggests that
Cpd1p might also be involved in similar cellular defen-
Fig. 6 Resistance to singlet oxygen-generating photosensitizers of sive mechanisms against oxygen stress.
YPH98 strains overexpressing or not expressing CPD1 or SNQ2. Although Cpd1p homologues are usually referred to
MB and TB are methylene blue and toluidine blue respectively. as cytosolic enzymes, plants possess chloroplastic GR or
Strains and other details are described in Figs. 2 and 5. lc denotes THR (Creissen et al. 1992; Mouaheb et al. 1998). We
low copy number plasmid
found that Cpd1p has 1±3 strong hydrophobic segments
suggestive of transmembrane or membrane associated
tizer toxins. S. cerevisiae proved to be an appropriate domains. Further in silico analysis suggests that Cpd1p
system for our purposes. We identi®ed two genes, CPD1 has a 68.5% probability to be plasma membrane-asso-
and SNQ2, that upon over-expression mediate increased ciated. Similarly, all Cpd1p homologous reductases also
resistance to cercosporin and other singlet oxygen-gen- contain 1±3 strong hydrophobic segments located in
erating photosensitizers. In contrast, cpd1 and snq2 null regions analogous to Cpd1p. These putative membrane-
mutants did not show greatly enhanced sensitivity to associated regions correspond to short segments found
photosensitizers, an observation which isin agreement within the domains involved in FAD and NAD(P)H
with previous reports which have shown that yeast binding, and in the dimerization domain (see Fig. 3).
possesses multiple overlapping defense mechanisms The involvement of a putative membrane-associated
against oxidative stress (Muller 1996; Izawa et al. 1998). reductase (Cpd1p) and of a plasma membrane multidrug
While Snq2p is a known transporter, Cpd1p is a novel exporter (Snq2p) is in agreement with a number of
putative reductase. The fact that Snq2p was not identi- previous reports, addressing cellular oxidative resistance
®ed when using a high-copy genomic library might mechanism. First, it has been suggested, and in some
suggest that the library used did not contain an intact cases shown, that the action of exogenous cytotoxic
SNQ2 allele. photosensitizers is con®ned to plasma membrane loci,
The direct role of Snq2p in multidrug and steroid indicating very limited migration within the cellular en-
export has been previously established (Servos et al. vironment (Kessel and Woodburn 1995). Thus, it should
1993; Decottignies et al. 1995; Mahe et al. 1996). Its be expected that proteins neutralizing cercosporin and
topology as a plasma membrane protein has also been other photosensitizers might be located in the plasma
con®rmed (Decottignies et al. 1995). In contrast, no membrane. Second, cercosporin self-resistance of Cer-
studies have previously addressed questions concerning cospora species appears to rely on multiple defense
the structure, topology and function of Cpd1p. The mechanisms that involve, among others, a putative
possible involvement of Cpd1p in the detoxi®cation of transporter (Cfp1p; Callahan et al. 1999) and the
lipophilic photosensitizers and the in silico analysis reduction of cercosporin molecules (Daub et al. 1992,
 135

2000). Third, MDR exporters have been implicated in Bilski P, Li MY, Ehrenshaft M, Daub ME, Chignell CF (2000)
the detoxi®cation of various lipid soluble toxic com- Vitamin B6 (pyridoxine) and its derivatives are ecient singlet
oxygen quenchers and potential fungal antioxidants. Photo-
pounds including photosensitizers, which they recognize chem Photobiol 71:129±34
in the cytoplasmic lea¯et of the bilayer and may function Bissinger PH, Kuchler K (1994) Molecular cloning and expression
as ¯ippases (Ruetz and Gros 1994; Kessel and Wood- of the Saccharomyces cerevisiae STS1 gene product. J Biol
burn 1995). Fourth, mammalian cellular lines develop- Chem 269:4180±4186
Callahan TM, Rose MS, Meadle MJ, Ehrenshaft M, Upchurch RG
ing resistance to lipid-soluble drugs (antifolate, (1999) CFP, the putative Cercosporin transporter of Cercospora
trimetrexate, piritrexim, etc.) in culture have been shown kikuchii, is required for wild type Cercosporin production, re-
to develop two phenotypic changes independently: either sistance and virulence on soybean. Mol Plant Microbe Interact
over-expressionof a multidrug resistance exporter 12:901±910
(Mdr1p), or over-expression of a transmembrane Chung KR, Jenns AE, Ehrenshaft M, Daub ME (1999) A novel
gene required for cercosporin toxin resistance in the fungus
reductase (dihydrofolate reductase; Assaraf et al. 1989; Cercospora nicotianae. Mol Gen Genet 262:382±389
Li et al. 1993). Creissen G, Edwards EA, Enard C, Wellburn A, Mullineaux P
Daub and colleagues have recently reported the iso- (1992) Molecular characterization of glutathione reductase
lation of two cercosporin resistance-related genes from cDNAs from pea (Pisum sativum L). Plant J 2:129±131
Daub ME (1982) Cercosporin, a photosensitizing toxin from
C. nicotianae. The ®rst, SOR1, encodes a novel pyri- Cercospora species. Phytopathology 72:370±374
doxine (vitamin B6) pathway gene (Ehrenshaft et al. Daub ME (1987) Resistance of fungi to the photosensitizing toxin
1998, 1999a). This observation suggests that pyridoxine cercosporin. Phytopathology 77:1515±1520
may be playing a novel biological role in possessing Daub ME, Briggs SP (1983) Changes in tobacco cell membrane
composition and structure caused by the fungal toxin cerco-
antioxidant activity (Ehrenshaft et al. 1999b; Bilski et al. sporin. Plant Physiol 71:763±766
2000). The other gene, termed CRG1, is required spe- Daub ME, Ehrenshaft M (1995) The photoactivated toxin
ci®cally for resistance to cercosporin and not for other cercosporin as a tool in fungal photobiology. Physiol Plant
photosensitizers (Chung et al. 1999). The biochemical 89:227±236
function of CRG1 remains unknown. These results, Daub ME, Leisman GB, Clark RA, Bowden EF (1992) Reductive
detoxi®cation as a mechanism of fungal resistance to singlet
along with those presented in this work, demonstrate a oxygen-generating photosensitizers. Proc Natl Acad Sci USA
rich source of microbial genes which can be used against 89:9588±9592
oxidative damage, plant disease and aging. Daub ME, Li M, Bilski P, Chignell CF (2000) Dihydrocercosporin
singlet oxygen production and subcellular localization: a pos-
Acknowledgements We would like to thank Drs G. Thireos, sible defense against cercosporin phototoxicity in Cercospora.
D. Alexandraki, D. Tzamarias, K. Kuchler, R. Egner, C. Sengstag, Photochem Photobiol 71:135±140
R. Gaber and L. Riles for providing us with the plasmids and Decottignies A, Kolaczkowski M, Balzi E, Go€eau A (1994) Sol-
strains used in this work. We are also grateful to Drs M. Daub and ubilization and characterization of the overexpressed PDR5
R. Upchurch for sharing with us unpublished data concerning multidrug resistance nucleotide triphosphatase of yeast. J Biol
clone cLE6, and to Mr. K. Damianakis for technical assistance. Chem 269:12797±12803
G.D., D.G. and P.V. were supported by a General Secretariat Decottignies A, Lambert L, Catty P, Degand H, Epping EA,
Research and Technology Grant awarded to N.P., F.D. was sup- Moye-Rowley WS, Balzi E, Go€eau A (1995) Identi®cation and
ported in part by scholarships from the Institute of Molecular characterization of SNQ2, a new multidrug ATP binding cas-
Biology and Biotechnology and the EPEAEK program of the sette transporter of the yeast plasma membrane. J Biol Chem
University of Crete. This work was ®nanced by a General 270:18150±18157
Secretariat Research and Technology Grant. Eggink G, Engel H, Vriend G, Terpstra P, Witholt B (1990)
Rubredoxin reductase of Pseudomonas oleovorans structural
relationship to other ¯avoprotein oxidoreductases based on one
NAD and two FAD ®ngerprints. J Mol Biol 212:135±142
Ehrenhofer-Murray AE, Wurgler FF, Sengstag C (1994) The
References Saccharomyces cerevisiae SGE1 gene product: a novel drug-
resistance protein within the major facilitator superfamily. Mol
Adams A, Gottschling DE, Kaiser CA, Stearns T (1997) Methods Gen Genet 244:287±294
in yeast genetics. Cold Spring Harbor Laboratory Press, Cold Ehrenhofer-Murray AE, Seitz MU, Sengstag C (1998) The Sge1
Spring Harbor, N.Y. protein of Saccharomyces cerevisiae is a membrane-associated
Andre B (1995) An overview of membrane transport proteins in multidrug transporter .Yeast 14:49±65
Saccharomyces cerevisiae. Yeast 11:1575±1611 Ehrenshaft M, Jenns AE, Daub ME (1995) Targeted gene disrup-
Arrick BA, Grith OW, Cerami A (1981) Inhibition of glutathione tion of carotenoid biosynthesis in Cercospora nicotianae reveals
synthesis as a chemotherapeutic strategy for trypanosomiasis. no role of carotenoids in photosensitizer resistance. Mol Plant
J Exp Med 153:720±725 Microbe Interact 8:569±575
Assaraf YG, Molina A, Schimke RT (1989) Sequential ampli®ca- Ehrenshaft M, Jenns AE, Chung KR, Daub ME (1998) SOR1, a
tion of dihydrofolate reductase and multidrug resistance genes gene required for photosensitizer and singlet oxygen resistance
in Chinese hamster ovary cells selected for stepwise resistance to in Cercospora fungi, is highly conserved in divergent organisms.
the lipid-soluble antifolate trimetrexate. J Biol Chem Mol Cell 1:603±609
264:18326±18334 Ehrenshaft M, Chung KR, Jenns AE, Daub ME (1999a) Func-
Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, tional characterization of SOR1, a gene required for resistance
Smith JA, Struhl K (1994) Current protocols in molecular to photosensitizing toxins in the fungus Cercospora nicotianae.
biology. Wiley, New York Curr Genet 34:478±485
Balzi E, Wang M, Leterme S, Van Dyck L, Go€eau A (1994) Ehrenshaft M, Bilski P, Li MY, Chignell CF, Daub, ME (1999b)
PDR5, a novel yeast multidrug resistance conferring trans- A highly conserved sequence is a novel gene involved in de
porter controlled by the transcription regulator PDR1. J Biol novo vitamin B6 biosynthesis. Proc Natl Acad Sci USA
Chem 269:2206±2214 96:9374±9378
136

Gaber RF, Kielland-Brandt MC, Fink GR (1990) HOL1 muta- heterologous expression in the yeast Saccharomyces cerevisiae.
tions confer novel ion transport in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 95:3312±3317
Mol Cell Biol 10:643±652 Muller EG (1991) Thioredoxin de®ciency in yeast prolongs S phase
Go€eau A, Park J, Paulsen IT, Jaunniaux J-L, Dinh T, Mordant P, and shortens the G1 interval of the cell cycle. J Biol Chem
Saier MH (1997) Multidrug-resistant transport proteins in 266:9194±9202
yeast: complete inventory and phylogenetic characterization of Muller EG (1996) A glutathione reductase mutant of yeast accu-
yeast open reading frames within the major facilitator super- mulates high levels of oxidized glutathione and requires thior-
family. Yeast 13:43±54 edoxin for growth. Mol Biol Cell 7:1805±1813
Haase E, Servos J, Brendel M (1992) Isolation and characterization Nitiss J, Wang JC (1988) DNA topoisomerase-targeting antitumor
of additional genes in¯uencing resistance to various mutagens drugs can be studied in yeast. Proc Natl Acad Sci USA 85:7501±
in the yeast Saccharomyces cerevisiae. Curr Genet 21:319±324 7505
Halliwell B, Gutteridge JM (1990) Role of free radicals and Paulsen IT, Brown MH, Skurray RA (1996) Proton-dependent
catalytic metal ions in human disease: an overview. Methods multidrug e‚ux systems. Microbiol Rev 60:575±608
Enzymol 186:1±85 Robzyk K, Kassir Y (1992) A simple and highly ecient procedure
Hirata D, Yano K, Miyahara K, Miyakawa T (1994) Sacchar- for rescuing autonomous plasmids from yeast. Nucleic Acids
omyces cerevisiae YDR1, which encodes a member of the ATP- Res 20:3790
binding cassette (ABC) superfamily, is required for multidrug Ruetz S, Gros P (1994) A mechanism for P-glycoprotein action in
resistance. Curr Genet 26:285±294 multidrug resistance: are we there yet? Trends Pharmacol Sci
Holmgren A (1985) Thioredoxin. Annu Rev Biochem 54:237±271 15:260±263
Ito T (1981) Dye binding and photodynamic action. Photochem Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning, a
Photobiol 33:947±955 laboratory manual, 2nd edn. Cold Spring Harbor Laboratory
Izawa S, Maeda K, Miki T, Mano J, Inoue Y, Kimura A (1998) Press, Cold Spring Harbor, N.Y
Importance of glucose-6-phosphate dehydrogenase in the Sanglard D, Kuchler K, Ischer F, Pagani JL, Monod M, Bille J
adaptive response to hydrogen peroxide in Saccharomyces (1995) Mechanisms of resistance to azole antifungal agents in
cerevisiae. Biochem J 330:811±817 Candida albicans isolates from AIDS patients involve speci®c
Jenns AE, Scott DL, Bowden EF, Daub ME (1995) Isolation of multidrug transporters. Antimicrob Agents Chemother
mutants of the fungus Cercospora nicotianae altered in their 39:2378±2386
response to singlet-oxygen-generating photosensitizers. Photo- Schutz M, Shahak Y, Padan E, Hauska G (1997) Sul®de-quinone
chem Photobiol 61:488±493 reductase from Rhodobacter capsulatus. Puri®cation, cloning
Kessel D, Woodburn K (1995) Selective photodynamic inactivation and expression. J Biol Chem 272:9890±9894
of a multidrug transporter by a cationic photosensitising agent. Servos J, Haase E, Brendel M (1993) Gene SNQ2 of Saccharomyces
Br J Cancer 71:306±310 cerevisiae, which confers resistance to 4-nitroquinoline-N-oxide
Knox JP, Dodge AD (1985) Singlet oxygen and plants. Phyto- and other chemicals, encodes a 169 kDa protein homologous to
chemistry 24:889±896 ATP-dependent permeases. Mol Gen Genet 236:214±218
Kolaczkowski M, Rest M van der, Cybularz-Kolaczkowska A, Shames SL, Fairlamb AH, Cerami A, Walsh CT (1986) Puri®-
Soumillion J-P, Konings WN, Go€eau A (1996) Anticancer cation and characterization of trypanothione reductase from
drugs, ionophoric peptides, and steroids as substrates of the yeast Crithidia fasciculata, a newly discovered member of the family
multidrug transporter Pdr5p. J Biol Chem 271:31543±31548 of disul®de-containing ¯avoprotein reductases. Biochemistry
Kuriyan J, Krishna TSR, Wong L, Guenther B, Pahler A, Williams 25:3519±3526
CH Jr, Model P (1991) Convergent evolution of similar Sikorski RS, Hieter P (1989) A system of shuttle vectors and yeast
function in two structurally divergent enzymes. Nature 352: host strains designed for ecient manipulation of DNA in
172±174 Saccharomyces cerevisiae. Genetics 122:19±27
Leishman GB, Daub, ME (1992) Singlet oxygen yields, optical Sollod CC, Jenns AE, Daub ME (1992) Cell surface redox potential
properties, and phototoxicity of reduced derivatives of the as a mechanism of defense against photosensitizers in fungi.
photosensitizer cercosporin. Photochem Photobiol 55:373±379 Appl Environ Microbiol 58:444±449
Li XK, Kobayashi H, Holland JF, Ohnuma T (1993) Expression Spikes JD (1989) Photosensitization. In: Smith KC (ed) The science
of dihydrofolate reductase and multidrug resistance genes in of photobiology, 2nd edn. Plenum Press, New York, pp 79±110
trimetrexate-resistant human leukemia cell lines. Leuk Res Tartaglia LA, Storz G, Brodsky MH, Lai A, Ames BN (1990)
17:483±490 Alkyl hydroxide reductase from Salmonella typhimurium. J Biol
Mahe Y, Lemoine Y, Kuchler K (1996) The ATP binding cassette Chem 265:10535±10540
transporters Pdr5 and Snq2 of Saccharomyces cerevisiae can Upchurch RG, Walker DC, Rollins JA, Ehrenshaft M, Daub ME
mediate transport of steroids in vivo. J Biol Chem 27:25167± (1991) Mutants of Cercospora kikuchii altered in cercosporin
25172 synthesis and pathogenesis. Appl Environ Microbiol 57:2940±
Meister A, Anderson ME (1983) Glutathione. Annu Rev Biochem 2945
52:711±760 Wierenga RK, Terpstra P, Hol WGJ (1986) Prediction of the
Mehdy MC, Sharma YK, Sathasivan K, Bays NW (1996) The role occurrence of the ADP-binding bab-fold in proteins, using an
of activated oxygen species in plant disease resistance. Physiol amino acid sequence ®ngerprint. J Mol Biol 187:101±107
Plant 98:365±374 Williamson JD, Scandallios JG (1993) Response of the maize
Meshnick SR, Blobstein SH, Grady RW, Cerami A (1978) An catalases and superoxide dismutases to cercosporin-containing
approach to the development of new drugs for African tryp- fungal extracts: the pattern of catalase response in scutella is
anosomiasis. J Exp Med 148:569±579 stage speci®c. Physiol Plant 88:159±166
Mouaheb N, Thomas D, Verdoucq L, Monfort P, Meyer Y (1998) Young AJ (1991) The photoprotective role of carotenoids in higher
In vivo functional discrimination between plant thioredoxins by plants. Physiol Plant 83:702±708