(Received for publication, July 18, 1996, and in revised form, February 11, 1997)
From the Fre1p and Fre2p are ferric reductases which
account for the total plasma membrane associated activity, a
prerequisite for iron uptake, in Saccharomyces cerevisiae.
The two genes are transcriptionally induced by iron depletion. In this
communication, we provide evidence that Fre2p has also cupric reductase
activity, as has been previously shown for Fre1p (Hassett, R., and
Kosman, D.J. (1995) J. Biol. Chem. 270, 128-134).
Both Fre1p and Fre2p enzymes are functionally significant for copper
uptake, as monitored by the accumulation of the copper-regulated
CUP1 and CTR1 mRNAs in fre1 Iron and copper are elements that have many similar chemical
properties. Both are transition metals with several oxidation states,
have close atomic and ionic radii numbers, and have very similar
electronegativity. They are relatively homogeneously distributed on the
planet and have thus been integrated in many biochemical reactions
during evolution. Both are essential for life in almost every species
(bacteria, fungi, plants, mammals) (1, 2). Excess quantities of iron
and copper exert similar amount-dependent cytotoxicity,
both favoring the formation of hydroxyl radicals which are disastrous
to the cell (3). Iron presents a particular problem for its uptake,
since in the oxygen-containing atmosphere it is practically all in the
form of insoluble ferric hydroxides. Organisms have therefore developed
complex mechanisms of high fidelity and precision to achieve an
appropriate homeostasis of these two metals.
In Saccharomyces cerevisiae, two proteins of the plasma cell
membrane, Fre1p and Fre2p, reduce Fe(III) to Fe(II) in the proximal vicinity of the cell. The expression of both corresponding genes is
regulated by the environmental iron concentration by a negative feedback mechanism which takes place at the transcriptional level (4,
5). The coupled function of the cell surface Fet3p multicopper oxidase,
which catalyzes the conversion of Fe(II) to Fe(III) extracellularly (6), and the recently reported Ftr1p permease (7) are also required for
high affinity iron uptake. Thus, a link between iron and copper
metabolism was first noted by the isolation of the FET3 gene
in a scheme aiming to clone the ferrous transporter (8). Fet3p requires
copper to function, and therefore high affinity iron uptake requires
copper (8, 9). Mutations either in the high affinity copper transporter
gene CTR1 or in the CCC2 gene, encoding a member
of the family of P-type ATPases proposed to transport cytosolic copper
into the lumen of a secretory organelle, results in iron deficiency in
the cell (9, 10). A similar mechanism has been postulated in mammals
for the release of newly absorbed iron from intestine to blood
involving the plasma glycoprotein ceruloplasmin, a copper-binding
protein with ferrous oxidase activity (11, 12). Finally, the iron
regulated transcription of the yeast genes FRE1,
FRE2, FET3, CCC2, FTR1, and
FTH1 (FTR1 homologue of unknown function) is
affected by the Aft1p transcriptional activator which recognizes a
specific consensus sequence on their promoters (13, 14). Moreover,
expression of FRE1 and CTR1 mRNAs depends on
the nuclear protein Mac1p, which is involved in iron and copper
utilization (15).1
We have previously shown that although Fre1p and Fre2p have seemingly
redundant functions in S. cerevisiae, the two genes are
up-regulated by the absence of extracellular iron in a kinetically different way, implying that they are subject to distinct regulation (5). Since there is evidence that copper reduction might be a component
of copper uptake (16) and that Fre1p is also a copper-repressible cupric reductase (15-17), in this study we have investigated the participation of FRE1 and FRE2 gene products in
copper metabolism. Our data clearly point to a role of both activities
in copper uptake, although the two genes are differentially affected by the function of Mac1p under conditions of copper (and iron) depletion. Mac1p transactivating function is itself modulated by the availability of copper, being higher in its absence.
The yeast
strains used in this study are all derivatives of the S288C strain. The
wild type and fre1 The growth media used were SD (2% glucose, 0.67% yeast nitrogen base;
Difco) or SD supplemented either with 100 µM
bathophenanthroline disulfonic acid-Na2 salt (BPS) (Fe(II)
and Cu(II) chelator) or with 100 µM bathocuproine
disulfonic acid-Na2 salt (BCS) (Cu(I) chelator) (16, 19,
20). For all experiments described, cells were grown to saturation in
SD medium, subsequently resuspended in the same medium, and grown for
8-10 generations (exponential phase). They were then shifted to the
desired medium (SD, SDBCS, or SDBPS) at a concentration of 4-5 × 106 cells/ml. Aliquots were removed from each culture every
2 h and assayed for copper and iron reducing activities
simultaneously.
-The FRE2/lacZ and
FRE1/lacZ fusion plasmids were constructed starting from the
low copy number URA3- containing YCP50 vector carrying a
6.3-kilobase BamHI-SalI fragment containing the
Escherichia coli lacZ gene (except for the 8 amino-terminal
codons of The MAC1 gene was isolated from a YCP50 yeast genomic
library by the use of a synthetic oligonucleotide probe. The sequence of the oligonucleotide is 5 The sequences of the antisense CTR1, CUP1, and
EF1 End-labeling of these oligonucleotides with Samples of total
RNA and yeast DNA were prepared and blot-hybridized according to
standard procedures (22).
-For the FRE1
gene-specific probe, a PvuII-SacI genomic DNA
fragment containing the coding region of the FRE1 gene
( For the FRE2 gene-specific probe, an
MluI-AflII genomic DNA fragment containing the
coding region of the FRE2 gene ( For the HIS3 gene-specific probe, a HindIII
genomic DNA fragment covering the coding region of the HIS3
gene from +305 to +591 was subcloned into the HindIII site
of the phagemid described for the FRE1-specific probe. The
recombinant phagemid was linearized by restriction at the
MseI site (+312) and a 260-base probe was produced by
transcription with the T7 RNA polymerase.
105 cpm of each
antisense RNA probe were simultaneously hybridized with 25 µg of
total RNA in a 20-µl reaction mixture containing 75% formamide, 0.5 M NaCl, 10 mM Tris pH 7.5, and 1 mM
EDTA overnight. The mixture was diluted in 300 µl of a solution
containing 0.3 M NaCl and 5 mM EDTA and treated
with RNase A (30 µg/ml) at 20 °C for 30 min, followed by treatment
with proteinase K (300 µg/ml in 0.1% SDS) at 37 °C for 20 min.
The mixture was phenol/chloroform extracted, ethanol precipitated,
denatured at 92 °C for 5 min, and electrophoresed on a 40-cm, 6%
polyacrylamide/bisacrylamide (19:1), 8.3 M urea-containing
gel.
Ferric reduction activity was measured as
described previously (5). Exactly the same conditions were used for the
copper reduction assay, except that BCS and CuCl2 was used
instead of BPS and FeCl3. For the Cu(I) quantification, an
extinction coefficient of 12.25 mM It has been shown previously that
Fre1p ferric reductase is also cupric reductase induced by copper
depletion (16). In this report, we have examined Fre2p ferric reductase
for cupric reducing activity assayed in a fre1 Table I.
Fre1p and Fre2p induction under copper and iron depletion
Foundation for Research and
Technology-Hellas,
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
,
fre2
, and fre1
fre2
mutant
strains. However, only Fre1p activity is induced by copper depletion,
even in the presence of iron. This differential copper-dependent
regulation of Fre1p and Fre2p is exerted at the transcriptional level
of the two genes. We have shown that Mac1p, known to affect the basal
levels of FRE1 gene expression (Jungmann, J., Reins, H.-A.,
Lee, J., Romeo, A., Hassett, R., Kosman, D., and Jentsch, S. (1993)
EMBO J. 12, 5051-5056), accounts for both the
copper-dependent induction of FRE1 and
down-regulation of FRE2 gene. Finally, Mac1p
transcriptional activation function is itself modulated by the
availability of copper.
Yeast Strains, Media, and Growth Conditions
, fre2
, and
fre1
fre2
strains have been previously
described (5). For the MAC1 gene disruption in the yeast
genome, a 480-base pair StyI fragment (codons 41-201) was
replaced by the 1.1-kilobase HindIII fragment of the
URA3 gene on a MAC1 (
100 to 963)/pBluescript
recombinant phagemid. The resulted insertion fragment was excised and
used to transform the desired ura3-52 strains (18). Uracil
prototrophy was used for the selection of transformants. Transformants
able to grow on glycerol only in the presence of 50 µM
copper (15) were further confirmed for the MAC1 gene
deletion by DNA blot hybridization. The yeast strain used for
transcriptional induction and repression assays of the LexA-Mac1p
fusion proteins to lacZ reporters was L9FT5, a derivative of
FY105 (MAT
,
his3
::LexAopHIS3 leu2::PET56 ura3-52 trp1
63).
-galactosidase), the lacY gene, and a portion
of the lacA gene. The
977/+3 and
930/+3 promoter
fragments of FRE1 and FRE2 genes, respectively, produced by polymerase chain reaction were cloned in the
BamHI site of the described plasmid in fusion to the
-galactosidase open reading frame.
-GCCCTCTGATGCACGATGCACACGCA-3
. For the
construction of the LexA-Mac1p fusion proteins, the MAC1
gene sequence from +4 to +1249 nucleotides was synthesized by
polymerase chain reaction and subcloned into the
SmaI-KpnI sites of plasmid YCp91, a derivative of
pRS314 (TRP1) containing the ADH1 promoter and 5
untranslated region followed by an ATG codon and sequences encoding the
SV40 nuclear localization signal, the HA1 epitope from influenza virus,
and the LexA binding domain (21). The oligonucleotide sequences used
for the polymerase chain reaction are 5
-TCCCCGGGATAATATTTAATGGGAACA-3
and 5
-GGGGTACCTGAAGTGGTGGCATCGCTTA-3
. The reporter plasmids used for
the transactivation and repressor assays were correspondingly pJK103,
in which four LexA operators upstream of the GAL1
TATA element control the lacZ gene transcription (21, 22),
and pJK1621, in which four LexA operators, two upstream activation sequences, and TATA element of the CYC1 gene
control the lacZ gene transcription (23, 24).
synthetic oligonucleotides used as probes in the RNA
blot hybridization analyses are 5
-CCTGCAACTTGGAATTCTCAAGGATGTC-3
,
5
-GTTACCGCAGGGGCATTTGTCGTCGTC-3
, and
5
-CCAGAATCGACATGACCGATAACGAC-3
, respectively.
-[32P]ATP
was performed by DNA kinase (22). The described oligonucleotides were synthesized by the Institute of Molecular Biology and Biotechnology Microchemistry group (Heraklion, Crete, Greece).
680 to +2280) was subcloned into the EcoRV site of the
pBluescript II KS (+/-) phagemid (Stratagene). The recombinant plasmid
was linearized by restriction at the SpeI site on the
FRE1 gene (+1893) and a 444-base (including polylinker
sequences) single-stranded homogeneously radiolabeled antisense RNA
probe was synthesized by the T3 RNA polymerase in the presence of
-[32P]UTP (21). From the 444-base FRE1
probe, a 298-base fragment was protected, which defined the 3
end of
the RNA 135 bases downstream of the translation termination. A
protected doublet was observed with this probe due to three consecutive
thymidines preceding this point, which created an unstable RNase
digestion-prone A/T-rich region.
2041 to +2340) was
subcloned into the EcoRV site of the phagemid described for
the FRE1-specific probe. The recombinant plasmid was
linearized by restriction at the AhaII site on the
FRE2 gene (+1682) and a 712-base probe was produced by the
same procedure as for the FRE1-specific probe. From the
712-base FRE2 probe, 654 bases were protected,
i.e. the size of the whole probe omitting the polylinker sequences (the FRE2 mRNA 3
is downstream the
AflII site used for subcloning). A protected doublet was
observed with this probe due to a region of seven consecutive
thymidines preceding the AflII site.
1
cm
1 was used to measure the Cu(I)BCS complex at 482 nm
(A482). The extinction coefficient used for
Fe(II)BPS at 520 nm (A520) was 22.39 mM
1 cm
1.
-Galactosidase Assays
-Galactosidase activity assays
(22) were performed following 7 h of growth in the desired
medium.
Fre1p and Fre2p Both Are Cupric Reductases but Only Fre1p Is
Induced by Copper Deprivation
strain
(5). We have first used iron depletion culture conditions known to
induce Fre2p (5) to obtain measurable amounts of the enzyme. For
comparison, we have assayed three additional strains, wild type,
fre2
(Fre1p activity), and
fre1
fre2
, in parallel according to the
conditions described under "Experimental Procedures." The results
presented in Table I show that Fre2p as well as Fre1p
have cupric reductase activity.
Medium
Strain
Cupric reductase
units
Ferric reductase units
4 h
10 h
4
h
10 h
SD
WT
5.3
± 0.8
6.3 ± 2.0
4.5 ± 1.3
5.5 ± 2.7
fre1
5.0 ± 1.8
4.8 ± 0.4
1.2
± 1.0
1.4 ± 0.9
fre2
4.2
± 1.0
4.6 ± 0.5
6.1 ± 3.5
4.2 ± 1.9
fre1
fre2
4.7 ± 2.0
3.6
± 0.3
0.8 ± 1.4
1.4 ± 1.1
SDBCS
WT
38.5 ± 4.5
18.8
± 1.8
102.5 ± 26.1
48.8
± 7.1
fre1
4.2 ± 0.0
5.3
± 1.3
1.9 ± 1.8
3.4 ± 1.8
fre2
35.5 ± 5.1
15.8
± 1.8
104.6 ± 15.9
45.8
± 4.8
fre1
fre2
4.2
± 0.4
4.7 ± 0.6
2.0 ± 1.7
3.2 ± 1.0
SDBPS
WT
40.2 ± 3.8
47.0
± 2.0
120.7 ± 35.0
91.2
± 31.8
fre1
4.4
± 2.7
32.0 ± 1.1
1.4 ± 1.4
53.6
± 17.0
fre2
31.8
± 7.9
12.0 ± 4.1
103.7
± 20.9
33.6 ± 18.2
fre1
fre2
5.2 ± 0.7
3.8
± 0.1
1.8 ± 1.6
1.8 ± 1.3
Since we had previously observed that the two enzymes follow
kinetically distinct induction by iron deprivation (5), we have assayed
the same strains under conditions of copper depletion during 12 h
of exponential growth. Our results (Table I) show that Fre1p was
induced under copper depletion at similar levels as by iron depletion.
In contrast, Fre2p was not detectably modulated by copper depletion.
Fre2p activity was not induced even in the presence of the reducing
agent sodium ascorbate (up to 500 µM), excluding the
possibility of inefficient retention of the Cu(II) found in the minimal
medium by the Cu(I)-specific chelator BCS (data not shown)). We have
obtained the same results (activity observed only in the
fre2 strain) by using medium reconstituted from its
ingredients with the omission of copper (data not shown).
Data in Table I show residual copper-reducing activity in the
fre1fre2
strain which was not affected by
metal depletion and accounted for about 80% of the activity in
noninduction conditions. We have not detected similar residual ferric
reducing activity in this fre1
fre2
strain
(5) (Table I).
Although Fre2p did not show induction within 12 h in
copper-depleted fre1 cultures, this was achieved later
between 12 and 15 h in repeatedly diluted copper-depleting medium
and remained relatively constant during the exponential growth of the
cells (Fig. 1). We suggest that according to the
proposed model of iron-copper connection (8, 9) prolonged depletion of
copper intracellularly created also iron depletion, which in turn
induced Fre2p activity.
Copper Reduction by Fre1p and Fre2p Cupric Reductases Facilitates Copper Uptake
-While our results showed that both Fre1p and Fre2p
are cupric reductases, they do not answer the question of whether
copper reduction is necessary for copper uptake. Assuming that Fre1p and Fre2p are indeed important for copper uptake, this process should
be impaired in a fre1fre2
strain. We have
monitored copper entrance into the cell and its utilization by
following the expression of two different copper responsive genes,
CUP1 and CTR1. The regulation of these genes by
copper follows distinct pathways. CUP1 transcript encoding a
metallothionein is induced by the transcription factor Ace1p when
copper concentration increases in the cell (25, 26), whereas
CTR1 transcript encoding the copper transporter is induced by copper depletion and is not affected by the Ace1p factor (27). As
shown in Fig. 2A, high levels of
CTR1 mRNA were detected in fre1
fre2
cells grown in minimal medium
(SD), which indicated that the mutant cells had lower copper uptake
capacity than the wild type cells. Nonetheless, addition of copper to a
concentration of 0.5 µM completely repressed
CTR1 mRNA levels in both strains, indicating that the
metal could enter also the fre1
fre2
cells. No significant difference was detected in the accumulation of CUP1 mRNA between the two strains, possibly because
CTR1 gene responded to lower levels of copper than
CUP1 gene (27).
When cultures grew under Fe(II) and Cu(II) retention conditions (SDBPS)
when both reductases were induced, prominent differences in the
accumulation of both CTR1 and CUP1 mRNAs were
obtained between fre1fre2
and wild type
strain. One hundredfold higher CuSO4 concentration was
necessary to diminish CTR1 mRNA levels in the mutant
cells. This observation demonstrated that the two reductases were
clearly required for copper uptake under conditions of strong retention
by chelators. As expected, when the two cultures grew under Cu(I)
depletion conditions (SDPBCS), the difference between the two strains
almost disappeared, since copper chelation was exerted after the step
of reduction. The small remaining difference might reflect a higher
Cu(I) concentration in the presence of reductases facilitating the
Cu(I) transporter to compete with the chelator.
To evaluate the contribution of each of the reductases for copper
uptake, we have similarly tested the singly disrupted strains fre1 and fre2
under conditions at which the
double disruptant and wild type cells had shown the most prominent
differences (SDBPS) (Fig. 2B). Comparison of the extent of
CTR1 mRNA repression in the four strains showed clearly
that both reductases participated in copper uptake. The similarity of
CTR1 mRNA patterns between fre2
and wild
type cells implied that Fre1p carried through the process almost as
well as both reductases. However, a clear-cut difference between the
fre1
and fre1
fre2
strains (a
5-fold CuSO4 concentration was required to repress
CTR1 mRNA in the doubly disrupted cells) was detected,
attributing to the Fre2p minor participation in copper uptake. The
detected CUP1 mRNA accumulation could not distinguish in
either case between the presence or absence of Fre2p, but clearly
pointed to the important contribution of Fre1p.
The differential induction profiles of the two
reductases in copper depleted cells led us to investigate the levels of
FRE1 and FRE2 mRNAs under these conditions.
Using an RNase A protection assay we were able to simultaneously detect
and quantitatively compare FRE1, FRE2, and
HIS3 (as an internal control) mRNAs in total RNA of a
wild type strain. Fig. 3A (left
panels) shows that FRE1 mRNA accumulated under
copper depletion and FRE2 mRNA was barely detectable
(only at very long film exposures) at all time points, as described for
the induction of the corresponding enzymatic activities. In parallel
experiments using iron-depleted cultures, FRE1 and
FRE2 mRNAs accumulated both following the
time-dependent induction observed for the corresponding
reductase activities (5) (Table I).
Measurements of -galactosidase activities driven by the
FRE1 or FRE2 promoters in wild type cells grown
under copper and iron depletion indicated clearly that the
FRE1 promoter was able to confer both copper- and
iron-regulated expression of
-galactosidase, whereas the
FRE2 promoter conferred only iron-regulated expression (Fig.
4).
The Function of Mac1p Accounts for the Differential Copper-dependent Regulation of FRE1 and FRE2 Gene Expression
The only known nuclear protein implicated in both
copper and iron metabolism, affecting basal expression of the
FRE1 gene is Mac1p (15). mac1 mutant cells suffer
from copper and, possibly, iron deficiency, since addition of copper or
iron (partly) can complement their phenotypes (15). Mac1p contains a
region homologous to the copper-dependent transcription
factor Ace1p (15). We have followed the response of FRE1 and
FRE2 genes to copper depletion in a MAC1
gene-deleted strain during 12 h of growth. The results shown in
Fig. 3A (right panels) revealed that
FRE2 mRNA was detectable in mac1 cells
grown in minimal medium (SD) and substantially induced following copper
depletion. In contrast, FRE1 mRNA basal levels were
significantly lower in the mac1
uninduced cells (SD medium) compared with the wild type levels and accumulated at very
reduced levels in the copper-depleted cells. Following iron depletion
of the mac1
strain, both FRE1 and
FRE2 mRNAs were induced. However, by comparison to the
induction patterns obtained in wild type cells, we found a temporal
shift of the FRE1 mRNA induction to later time points in
the exponentially growing mac1
culture, in contrast
to a temporal shift of the FRE2 mRNA accumulation to earlier time points (Fig. 3, A and B).
-Galactosidase activity levels produced by the FRE1/lacZ
and FRE2/lacZ plasmids in a mac1
strain grown
under copper or iron depletion conditions showed that the differential
accumulation of FRE1 and FRE2 mRNAs resulted
from the transcriptional response of the two genes (Fig. 4). Comparing
these results to the RNA accumulation profiles, we should note that at
7 h of growth (when the assays were performed) the accumulation of
-galactosidase was less in the mac1
strain compared
with the wild type strain when driven by the FRE1 promoter
and more when driven by the FRE2 promoter. The results of
reductase assays on copper- and iron-depleted fre1
mac1
and
fre2
mac1
cultures were in agreement with
the induction patterns of their transcripts (data not shown).
Therefore, under copper limitation, Mac1p function has a negative role
on FRE2 gene regulation, while it affects positively FRE1 gene. In the absence of Mac1p (in mac1
strain) under copper limitation, additional inducing factors might
exist for both genes, since FRE2 and FRE1
mRNA levels are higher than those detected in the noninducing
culture (minimal medium). Furthermore, as presented in the diagrams
shown in Fig. 3B, Mac1p seems to be involved in the
temporally differential expression of FRE1/FRE2 genes under iron limitation. These observations are elaborated further under "Discussion."
The
observed copper-dependent regulation of
FRE1/FRE2 genes affected by Mac1p function led us to
investigate the copper dependence of Mac1p protein function. First we
have tested the entire Mac1p coding sequence fused to the LexA binding
domain for its ability to activate transcription of a -galactosidase
reporter following binding to the LexA operator (21, 22). As shown in
Fig. 5, Mac1p protein was capable of transactivating the
-galactosidase reporter in cells grown in minimal medium. This
activity was significantly increased when cells were depleted from
copper, which suggests that the Mac1p protein was activated at these
conditions. Mac1p mRNA levels were unaffected under
copper-depleting conditions (data not shown). These results implicate a
direct effect of copper on Mac1p modulating its transcriptional
activation function. Testing of the LexA-Mac1p fusion for repression
function on a CYC1 promoter-driven lacZ
transcription (23, 24) showed inducing rather than repressing effect on
the constitutive expression of this reporter (Fig. 5), indicating that
Mac1p probably does not have repressing activity. This result is
relevant for the explanation given for its different role on the
expression of the FRE1/2 genes (see under
"Discussion").
Iron and copper are metals whose biological importance and common properties have been emphasized frequently. The cellular mechanisms by which living organisms exploit them to fulfill their needs and simultaneously protect themselves are now being unravelled (12). The externally directed enzymes Fre1p and Fre2p are situated at the "beginning" of the pathway that links extracellular to intracellular iron communication, leading to nuclear events that alter gene expression, as well as at the "end" of such a pathway, being themselves regulated by iron to control its influx into the cell. Although a common iron-related transactivator, Aft1p, has been identified (13, 14), the two reductase genes are differentially induced (Ref. 5 and this paper).
In this study we have examined the participation of the ferric reductases Fre1p and Fre2p in copper metabolism, and we have revealed elements of their metal-regulated expression. Previous reports have presented the ferrireductase plasma membrane activity (17) and Fre1p (16) as being repressed by copper. We have demonstrated that Fre2p is a cupric reductase, as is Fre1p, but only Fre1p is induced in copper starvation, accounting for 80-90% of the plasma membrane activity. The residual copper-specific reducing activity, which is not modulated by copper, probably corresponds to the Cu(II)-specific reductase, not repressible by iron, described by Hassett and Kosman (16). This reductase is possibly encoded by another FRE-homologous gene.2
We have further shown that the distinct response of the two cupric
reductases to copper depletion is reflected from the differential transcriptional regulation of the two genes. While FRE1 gene
was transcriptionally up-regulated, FRE2 gene did not seem
to respond to copper depletion. Furthermore, we have shown that Mac1p
nuclear protein contributed directly or indirectly to this differential response of the two reductase genes. FRE1 gene basal and
induced expression were highly dependent on its presence, in agreement with the notion that Mac1p is involved in the communication of the
copper starvation signal to the FRE1 gene. In contrast,
FRE2 gene basal and copper-induced expression were only
observed in the absence of Mac1p. An explanation for this phenomenon,
which needs to be proven, may lie in the following observations.
FRE1 promoter contains a pyrimidine-rich directly repeated
sequence, previously shown to mediate iron-regulated transcription of a reporter gene (4), distinct from the Aft1p binding consensus, which
could be responsible for the Mac1p effects, as discussed by
Yamaguchi-Iwai et al. (14). Similar pyrimidine-rich repeated sequences are also found on the FRE2 (5) and CTR1
(9) promoters in different orientations. CTR1 gene
expression is also dependent on the presence of Mac1p
(15).1 Only on the FRE2 promoter are these
sequences situated next to the TATA box between positions 54 and
80. (The transcription start site of FRE2 gene is at
position
5 from the first coding AUG1 and the potential
TATA box is at position
89.) If these sequences were indeed affected
by Mac1p, they might play different roles on the different promoters.
While affecting positively the expression of FRE1 and
CTR1 genes, Mac1p may result in steric hindrance of the
formation of the transcriptional complex on the FRE2
promoter (28). This hypothesis is also in agreement with our results indicating that Mac1p plays a major role in the temporally different induction of the two reductase genes by the use of the Fe/Cu chelator (BPS medium). We can conceive that upon metal deprivation Mac1p facilitates FRE1 expression and inhibits FRE2
expression. Accumulation of Aft1p and/or appearance of another
epistatic FRE2-specific inducer results in the
FRE2 late induction.
This scheme is in agreement with our additional findings on the role of Mac1p. First, Mac1p was able to transactivate a reporter gene (lacZ) when artificially brought to its promoter (by the LexA binding domain). Second, Mac1p did not show any repression function on a constitutively expressed reporter gene (CYC1(UAS)-lacZ). Third, an important finding that directly related Mac1p function to the cupric reductases was the fact that its transactivation function was modulated by the availability of copper, being increased in its absence. The precise role of Mac1p will be revealed by the detailed study of its functional domains as well as examination of the interacting elements of the transcriptional complexes in which it participates.
Our last finding concerns the role of the Fre1p/Fre2p cupric reductases on the uptake of copper. Asking the question whether copper reduction is necessary for its cellular uptake, the profit from such a reaction is not immediately obvious since Cu(II) ions, unlike Fe(III), are found in a soluble form in water-dominated environments. Evidence as to whether copper enters the cell as Cu(II) or Cu(I) was indirect. Several studies in plants have shown a concomitant increase in Fe(III) and Cu(II) reducing activities in response to depletion of these metals from the soil (29, 30). These observations could be explained either by introducing the notion of a general role of reductases in regulating the redox state of the plasma membrane which would affect channel gating and ion influx (17, 30) or a more specific role in copper uptake which would take into account environmental factors, e.g. the availability of Cu(II) ions which form very stable complexes with various environmental ligands. Hassett and Kosman (16) presented evidence for copper reductive assimilation in S. cerevisiae, since Pt inhibition (Fre1p inhibitor) of copper uptake was relieved when the reducing agent ascorbate was included in the uptake assay. Very recently, Knight et al. showed that FRE1 gene is necessary for the Ctr1p and Ctr3p function in 64Cu uptake (31).
We have obtained evidence for the importance of reduction for copper
uptake by using the CTR1 and CUP1 mRNAs as
reporters to monitor entrance and utilization of copper in the cell. We have demonstrated that Fre1p/Fre2p activities assure higher
intracellular copper levels, as shown by the increased CTR1
mRNA quantities in the fre1fre2
strain
compared with wild type cells. This activity was not an absolute
prerequisite for copper uptake, possibly due to the entrance of Cu(II)
ions or to Cu(I) produced by the residual plasma membrane
copper-reducing activity mentioned above. Our results provide evidence
that mostly Fre1p but also Fre2p have a major contribution to the
uptake of copper when we used a chelator capturing its oxidized form
(BPS). This contribution might be very useful for the organism, since
there are probably many chelators of Cu(II) in the ecosystem of
S. cerevisiae (e.g. citrate in fruits). The
involvement of the two ferric reductases in copper and iron metabolism
appears distinct and specific, particularly since Fre1p and Fre2p do
not respond to molybdenum, zinc, or manganese deprivation.1
In reference to the iron-copper connection already discussed, it seems
inevitable that ferric reductases have evolved to participate also in
copper metabolism. Since iron cannot enter the cells in conditions of
copper starvation, induction of the ferric reductases by iron depletion
would not be of any use if copper entrance was not facilitated
simultaneously. On the other hand, when copper is depleted, Fre1p is
first induced, able to assume most part of copper import into the
cells, and if, for some reason, copper continues to be rare, Fre2p
comes to its aid, being induced much later indirectly because of iron
deprivation. A clear contribution of Fre2p was seen both under
prolonged copper starvation and in the mac1
strain. Our
results show clearly that reduction of copper facilitates its entrance
into the cells. We propose that Fre1p and Fre2p, although not
indispensable for copper uptake, are actively and catalytically
participating in that process, especially under natural conditions.
In conclusion, the Fre1p/Fre2p system, involving two genes distally related in primary structure (5) functioning in close collaboration for iron and copper handling but with clearly differentiated controlling mediators, is a very promising experimental model in revealing "fine tuning" mechanisms that have evolved in living organisms to assure profitable metal homeostasis. It is interesting to note that, as revealed by the completed sequence of the S. cerevisiae genome,2 six additional open reading frames share similarities with the Fre1p and Fre2p sequences and possibly are involved in other specific membrane reductase activity functions.
We thank George Thireos and Dimitris Tzamarias for helpful suggestions; Apostolis Klinakis, Alexis Zafiropoulos, Thanassis Tartas, and Alexandra Boutla for help with plasmid and strain constructions; Georgia Houlaki for artistic assistance; and Lila Kalogeraki for photographic assistance.