close

Вход

Забыли?

вход по аккаунту

?

744

код для вставкиСкачать

. 13: 621–637 (1997)
The AFT1 Transcriptional Factor is Differentially
Required for Expression of High-Affinity Iron Uptake
Genes in Saccharomyces cerevisiae
CELIA CASAS1, MARTIu ALDEA1, CARME ESPINET1, CARME GALLEGO1, ROSARIO GIL† AND
ENRIQUE HERRERO1*
1
Departament de Ciències Mèdiques Bàsiques, Facultat de Medicina, Universitat de Lleida, Rovira Roure 44,
25198-Lleida, Spain
Received 20 August 1996; accepted 1 December 1996
High-affinity iron uptake in Saccharomyces cerevisiae involves the extracytoplasmic reduction of ferric ions by FRE1
and FRE2 reductases. Ferrous ions are then transported across the plasma membrane through the FET3
oxidase-FTR1 permease complex. Expression of the high-affinity iron uptake genes is induced upon iron deprivation.
We demonstrate that AFT1 is differentially involved in such regulation. Aft1 protein is required for maintaining
detectable non-induced levels of FET3 expression and for induction of FRE2 in iron starvation conditions. On the
contrary, FRE1 mRNA induction is normal in the absence of Aft1, although the existence of AFT1 point mutations
causing constitutive expression of FRE1 (Yamaguchi-Iwai et al., EMBO J. 14: 1231–1239, 1995) indicates that Aft1
may also participate in FRE1 expression in a dispensable way. The alterations in the basal levels of expression of the
high-affinity iron uptake genes may explain why the AFT1 mutant is unable to grow on respirable carbon sources.
Overexpression of AFT1 leads to growth arrest at the G1 stage of the cell cycle. Aft1 is a transcriptional activator
that would be part of the different transcriptional complexes interacting with the promoter of the high-affinity iron
uptake genes. Aft1 displays phosphorylation modifications depending on the growth stage of the cells, and it might
link induction of genes for iron uptake to other metabolically dominant requirements for cell growth. ? 1997 by
John Wiley & Sons, Ltd.
Yeast 13: 621–637, 1997.
No. of Figures: 6. No. of Tables: 3.
No. of References: 62.
  — AFT1; transcriptional factor; iron uptake; phosphorylation; respiratory growth
INTRODUCTION
Iron is an essential element for most living cells. It
participates in redox reactions involved in such
diverse processes as DNA synthesis and respiratory electron transport. Cells have developed a
variety of strategies to assimilate the limiting
amounts of soluble iron present in the external
media [see Crichton (1991) for an extensive
review]. Since the excess of iron may be toxic for
the cell, the mechanism for its uptake must be
*Correspondence to: Enrique Herrero.
†Present address: Instituto de Agroquímica y Tecnología de
Alimentos, CSIC, 46100-Burjassot, Valencia, Spain.
Contract grant sponsor: DGICYT.
Contract grant sponsor: CIRIT.
Contract grant sponsor: Ajuntament de Lleida.
CCC 0749–503X/97/070621–17 $17.50
? 1997 by John Wiley & Sons Ltd
exquisitely regulated in response to the available
iron concentration.
Most bacteria secrete low molecular weight
compounds called siderophores that bind iron in
the ferric form and make it available for envelope
receptors that transport iron to the bacterial cytoplasm (Bagg and Neilands, 1987; Guerinot, 1994).
Saccharomyces cerevisiae cells do not contain a
siderophore-mediated iron uptake mechanism. At
low concentrations of the metal, yeast cells assimilate iron through a high-affinity uptake mechanism, the first step of which is the reduction of
external Fe(III) to Fe(II) by plasma membraneassociated ferric reductases (Lesuisse and Labbe,
1989; Dancis et al., 1990; Eide et al., 1992). Two
different genes, FRE1 and FRE2, are responsible
.   .
622
for most of the ferric reductase activity in S.
cerevisiae growing cells (Dancis et al., 1992;
Georgatsou and Alexandraki, 1994). Once external ferric iron has been converted into the ferrous
form, this is transported into the cytoplasm with
the participation of the plasma membrane ferric
oxidase product of FET3 (Askwith et al., 1994; De
Silva et al., 1995). It has been proposed that Fet3
forms a complex with the permease encoded by
FTR1 (Stearman et al., 1996). The fact that Fet3
requires copper for its oxidase activity (Askwith et
al., 1994) explains the requirement of the CTR1
copper transporter for high-affinity iron assimilation (Dancis et al., 1994). The product of FRE1
also accounts for part of the copper reductase
activity involved in copper uptake in S. cerevisiae
(Hassett and Kosman, 1995), thus confirming that
iron and copper assimilation functions are closely
related in yeast.
Besides the high-affinity system, yeast cells possess a low-affinity mechanism for iron assimilation
that operates at high concentrations of the metal
(Dancis et al., 1992; Eide et al., 1992) and is
mediated by the product of the FET4 gene (Dix et
al., 1994). None of the elements participating in
the high-affinity system are involved in the lowaffinity one, and vice versa (Dix et al., 1994).
The expression of FRE1, FRE2, FET3 and
FTR1 (Dancis et al., 1992; Askwith et al., 1994;
Georgatsou and Alexandraki, 1994; YamaguchiIwai et al., 1995; Stearman et al., 1996) and the
levels of ferric reductase activity (Eide et al., 1992)
are both upregulated in iron deprivation conditions. FRE1 transcription is repressed by copper as
well (Hassett and Kosman, 1995). Although iron
concentration acts as a signal for expression of the
four genes of the high-affinity iron transport system, no clear picture of the transcription factor(s)
involved exists. FRE1 and FRE2 promoters contain elements resembling those recognized by the
Yap1 transcription factor (Harshman et al., 1988;
Dancis et al., 1992; Georgatsou and Alexandraki,
1994; Lesuisse and Labbe, 1995), although only
FRE2 responds to this factor involved in the
expression of stress tolerance genes (Georgatsou et
al., 1995). On the other hand, basal expression of
FRE1 depends on the Mac1 nuclear factor, also
involved in stress resistance (Jungmann et al.,
1993).
Recently, AFT1 has been identified as a gene
that mediates iron uptake in S. cerevisiae by regulating expression of FRE1, FRE2, FET3 and FTR1
in response to iron deprivation conditions
? 1997 by John Wiley & Sons, Ltd
(Yamaguchi-Iwai et al., 1995, 1996; Stearman et
al., 1996). A constitutively upregulated AFT1 mutant maintains high levels of ferric reductase activity and iron transport independently of external
iron concentration, as well as constitutive expression of the above-mentioned four genes. On the
other hand, another functionally altered AFT1
mutant is especially sensitive to iron deprivation
due to defects in the expression of the high-affinity
iron transport genes (Yamaguchi-Iwai et al., 1995).
The AFT1 product regulates expression of the
ferric reductases and iron transport genes through
binding to a consensus sequence present in the
promoters of these genes (Yamaguchi-Iwai et al.,
1996). In addition, AFT1 regulates expression of
two other iron-responsive genes (Yamaguchi-Iwai
et al., 1996), CCC2 (involved in intracellular copper transport) and FTH1 (a gene partially homologous to FTR1 with unknown function). The
sequence of AFT1 reveals that it is identical to
RCS1, previously characterized by our group by
the deregulation of cell size observed in diverse
mutant alleles of the gene (Gil et al., 1991). In this
work, we demonstrate that Aft1 is differentially
involved in expression of the genes coding for
the high-affinity iron uptake system and becomes
post-transcriptionally modified in response to
nutritional changes.
MATERIALS AND METHODS
Yeast strains, media and growth conditions
Yeast strains are listed in Table 1. Strains were
constructed using standard methods (Rose et al.,
1990) for diploid formation, sporulation and tetrad analysis. Yeast cells were transformed using
the lithium acetate procedure (Gietz et al., 1992).
YEPD medium (1% yeast extract, 2% peptone,
2% glucose) was employed for growth in rich
conditions. YEP medium lacks glucose. SD minimal medium (Rose et al., 1990) was supplemented
with the adequate carbon source (at 2% concentration except when a mixture of 2% ethanol plus
3% glycerol was employed) and the required amino
acids. When required, iron was supplemented as
ferrous sulphate at the concentrations indicated in
each experiment. For iron starvation conditions,
the iron chelator ferrozine (Sigma) was added to
the SD medium as indicated. Simultaneous starvation for iron and copper was achieved with
bathophenanthrolene disulfonate (BPS; Sigma).
For induction experiments, 2% galactose was

. 13: 621–637 (1997)
      
Table 1.
Strain
List of strains.
Genotype
1788
Source or reference and comments
MATa/MATá; isogenic diploid for leu2–3, 112
ura3–52 trp1–1 his4 can1r
DBY747 MATa leu2–3, 112 ura3–52 trp1–289 his3-Ä1 can1r
HR125
MATa leu2–3, 112 ura3–52 trp1–1 his3–532 his4
INVSc2
MATá ura3–167 his3-Ä200 GAL2
OL1
MATá leu2–3, 112 ura3–251, 337 his3–11, 15
PEY101
MATa ade2–1 his3–11, 15 leu2–3, 112 trp1–1 ura3–1
can1r snf1–11::HIS3
SFY526
MATa leu2–3, 112 ura3–52 trp1–901 his3-Ä200
ade2–101 lys2–801 gal4–542 gal80–538
URA3::GAL-lacZ can1r
TCA41–1 MATa leu2–3, 112 ura3–52 trp1–1 his3–532 his4
cyr1::URA3 cam
W303–1A MATa ade2–1 his3–11, 15 leu2–3, 112 trp1–1 ura3–1
can1r
Y153
MATa leu2–3, 112 ura3–52 trp1–901 his3-Ä200
ade2–101 lys2–801 gal4–542 gal 80–538
URA3::GAL-lacZ LYS2::GAL-HIS3 can1r
CML12
MATá leu2–3, 112 ura3–251, 337 his3–11, 15 aft1-Ä1
CML72
MATa leu2 ura3 trp1 his3 GAL2
CML126
MATa leu2–3, 112 ura3–52 trp1–1 his4 can1r
aft1-Ä5::URA3
MATa leu2–3, 112 ura3–52 trp1–1 his4 can1r
CML128
623
added to cultures that contained 2% raffinose as
carbon source. Yeast cells were grown at 30)C with
shaking.
Determination of cellular parameters
Cell number was determined with a haemocytometer. DNA content distributions were obtained
by flow cytometry (Tyers and Futcher, 1993) with
an Epics XL (Coulter Co.). Cell sizes were
measured in cells viewed at 400-fold magnification
in an LSM 310 Carl Zeiss laser scanning microscope (transmission mode, 543 nm laser wavelength), using the apparatus software to convert
images into real distances. At least 100 cells were
measured per sample, and volumes were calculated
as previously described (Gil et al., 1991).
DNA manipulations
Standard DNA manipulations were performed
according to Sambrook et al. (1989) or Ausubel et
al. (1989). DNA fragments were isolated from
agarose gels using Qiaex (Diagen). Southern analysis and colony screenings were performed with
? 1997 by John Wiley & Sons, Ltd
Lee and Levin (1992)
D. Botstein
Hubble et al. (1993)
Invitrogene
M. Jacquet
From F. Estruch. SNF1 disruption
in the W303–1A background
Clontech
From HR125
Durfee et al. (1993)
This work. From OL1, by disruption of
AFT1 (Gil et al., 1991)
This work. OL1 genetic background with trp1
from DBY747 and GAL2 from INVSc2
This work. Segregant haploid, after introducing
the aft1-Ä5 disruption in diploid strain 1788
This work. Cosegregant with CML126
digoxigenin-labeled probes following the instructions provided by the manufacturer (Boehringer).
DNA sequencing was carried out by standard
methods (Sanger et al., 1977).
Plasmid isolation and construction
Isolation of pMV2 (containing a fragment of
AFT1, see Results) from a YEp13 genomic DNA
library (Nasmyth and Tatchell, 1980) has been
described previously (Gil et al., 1991). The complete AFT1 gene was obtained in pCM4, which
was isolated as an independent clone from the
same library.
Plasmid pCM28, which contains AFT1 under
the control of the GAL1 promoter, was obtained
by cloning a 2·6-kb DraI-BglII fragment from
pCM4 into the episomal vector pYES2 (Invitrogen) digested with SacI and BamHI. To obtain this
construct in a centromeric plasmid, a 2·6-kb KpnIXbaI fragment from pCM28 was cloned into YCpGAL digested with PstI (ends were made blunt
with Klenow fragment before ligation), resulting in
plasmid pCM49. YCpGAL (obtained from Alan

. 13: 621–637 (1997)
.   .
624
Boyd) contains the GAL1-GAL10/UAS region
cloned as an EcoRI-BamHI fragment in the
polylinker of YCplac22 (Gietz and Sugino, 1988).
Protein fusions of the last C-terminal 277
amino acids from Aft1 to protein A and
â-galactosidase were obtained by subcloning a
1·5-kb NaeI-EcoRI fragment from pMV23 (Gil et
al., 1991) into pAX12 (Zueco and Boyd, 1992)
and pEX12 (Kusters et al., 1989) digested with
SmaI-EcoRI, resulting in plasmids pCM2 and
pCM3 respectively.
Plasmid pCM86 contains AFT1 fused in frame
to the GAL4 DNA binding domain (bd) region. It
was constructed by opening pGBT9 [containing
GAL4(1–147) (Bartel et al., 1993)] with EcoRI, followed by Klenow treatment, dephosphorylation
and ligation to the EcoRI-KpnI blunt-ended 2·6-kb
fragment from pCM28 (this work) that comprises
the entire AFT1 gene. The ligation gives an in
frame fusion 5*-GAL4bd region-AFT1, as checked
by Western blotting with Aft1 antibodies.
For constructing pCM110, a SplI-BamHI fragment of 381 bp from pUHD15–1 (Gossen and
Bujard, 1992) was made blunt ended and cloned
into the single XmaI site of pGBT9. The resulting
plasmid codes for Gal4bd fused in frame to the
C-terminal 122 amino acids of the herpes simplex
virus transcriptional activator VP16 (Gossen and
Bujard, 1992). The hybrid protein has transcriptional activation properties on GAL1-directed
reporter genes (see below).
Construction of strains carrying AFT1 deletions
The aft1-Ä1 deletion was made following the
strategy described in Gil et al. (1991), and it was
introduced in the OL1 background to obtain strain
CML12.
The aft1-Ä5 allele lacks the promoter region and
most of the AFT1 open reading frame. A pCM4
partial digestion with HindIII produced a 3·7-kb
fragment containing the whole AFT1 open reading
frame. This fragment was made blunt ended, and
subcloned into the SmaI and HincII sites of pBluescript SK+, resulting in pCM22. From this plasmid, the EcoRI-HindIII fragment (spanning from
342 bp upstream from the initiation codon to the
HindIII internal to AFT1) was substituted by a
URA3 cassette (flanked on both sides by translation stop codons in the three open reading frames)
from YDp-U (Berben et al., 1991). The plasmid
thus obtained, pCM24, was digested with BamHI
plus KpnI to generate a linear fragment that
? 1997 by John Wiley & Sons, Ltd
includes the deleted version of AFT1, which was
used to replace the chromosomal wild-type copy
by one-step replacement (Rothstein, 1983). The
final chromosomal construct was checked by
Southern blot analysis.
Northern blot analysis
Total RNA was prepared from yeast cells and
10 ìg per sample were run in low-formaldehyde
gels as described by Ausubel et al. (1989). Transfer
to Nylon+ membranes (Boehringer) was performed with a VacuBlot (Pharmacia) following
the instructions provided by the manufacturer
with some modifications regarding gel treatments
(15 min with distilled water, twice for 15 min
with 0·1 -NaOH, 15 min with 0·1 -Tris–HCl pH
7·5) and transfer conditions (90 min with distilled
water). RNA was crosslinked to the membrane
with a Stratalinker (120 000 ìJ/cm2, Stratagene)
and membranes were washed twice for 15 min
at 65)C with 20 m-PO4H3 pH 7·2, 1 m-EDTA,
1% sodium dodecyl sulfate (SDS) to remove
formaldehyde traces.
Hybridization and chemiluminescent detection
steps were performed as described by Engler-Blum
et al. (1993). Digoxigenin-labeled probes were prepared according to the instructions provided by
the manufacturer (Boehringer). DNA probes were
as follows: a 873-bp BamHI-HindIII fragment
internal to ICL1 (Fernández et al., 1992), the
oligonucleotide 5*-TGGTAGCCTTAACGACTG
CGCTA from bp 687 to 709 of the ADH2 open
reading frame (Russell et al., 1983), a 2·6-kb
KpnI-EcoRI fragment from pCM28 for AFT1, a
1460 bp EcoRI-BstEII fragment internal to FRE1
(Dancis et al., 1992), a 1545 bp EcoRI-NcoI fragment spanning 5* sequences and part of the FRE2
open reading frame (Georgatsou and Alexandraki,
1994), an internal 724 bp EcoRI fragment from
open reading frame AOB629 (Casamayor et al.,
1995), a polymerase chain reaction (PCR)amplified fragment from the first to the last nucleotide of the FET3 open reading frame (De Silva et
al., 1995), a PCR-amplified fragment from nucleotide +2 to nucleotide +1650 (Dix et al., 1994) of
FET4, and a 1·6-kb HindIII-BamHI fragment
from pYA301 (Gollwitz and Suves, 1980) for
ACT1.
Immunological techniques and protein analysis
The protein A and â-galactosidase fusions to the
last C-terminal 277 amino acids of Aft1 produced

. 13: 621–637 (1997)
      
by plasmids pCM2 and pCM3 respectively, were
used to obtain immunopurified antibodies from
crude rabbit polyclonal antiserum, as described by
Zueco and Boyd (1992). For Western blots, total
protein extracts from yeast cells were prepared as
described by Ausubel et al. (1989), or alternatively
using the cracking buffer described by Printen and
Sprague (1994). Where indicated, high molecular
weight proteins were resolved in 6% acrylamide–
0·2% bisacrylamide–SDS running gels adjusted to
pH 8·2 (Makowski and Ramsby, 1993). Protein
transfer to PVDF membranes (Millipore) was performed with a semi-dry system (Pharmacia), and
proteins were immunodetected by using the ECL
system (Amersham) as recommended by the
manufacturers.
In order to immunoprecipitate Aft1, lysates
from 100 optical density (600 nm) cell units were
prepared by vortexing cell suspensions in 100 ìl of
cold lysis buffer (200 m-Tris–HCl pH 8, 0·5 mEDTA,
1 m-dithiothreitol,
0·1 m-phenylmethylsulfonyl fluoride, 0·1 m benzamide, 2 ìg/
ml pepstatin A, 1 ìg/ml leupeptin, 0·1% Triton
X-100) in the presence of glass beads. Cell debris
was pelleted at 12 000 g for 15 min at 4)C. A
volume of 50 ìl of supernatant was added to
0·5 ml of cold binding buffer (20 m-Tris–HCl pH
8, 125 m-NaCl, 0·1% Triton X-100) and 10 ìl of
immunopurified anti-Aft1 antibody. After 1 h on
ice, 25 ìl of protein A–Sepharose (Sigma 1:1 in
binding buffer) was added and the mixture
was incubated at 4)C while rocking. Immune
complexes were washed twice with 0·5 ml cold
RIPA buffer (50 m-Tris–HCl pH 8, 500 mNaCl, 1% Triton X-100, 0·1% SDS, 0·1% sodium
deoxycolate) and once with 0·5 ml cold binding
buffer. Samples were boiled with 25 ìl of 2#gel
sample buffer and centrifuged, and the supernatants were subjected to SDS–polyacrylamide gel
electrophoresis.
For alkaline phosphatase treatment of Aft1,
immunoprecipitates from 25 optical density (600
nm) cell units were obtained as before, washed
once in CIP buffer (Ausubel et al., 1989) and
resuspended in 25 ìl of CIP buffer containing 0·1%
SDS. Samples were boiled for 2 min and 2 units of
calf intestine alkaline phosphatase (Boehringer)
were added. In some experiments a mixture of
10 m-p-nitrophenyl phosphate, 50 m-glycerophosphate and 50 m-sodium fluoride (final concentrations) was added to inhibit enzyme activity.
Samples were incubated for 90 min at 37)C, and
reactions were stopped by addition of 25 ìl of
? 1997 by John Wiley & Sons, Ltd
625
2#SSR buffer (4% SDS, 10% sucrose, 0·01%
bromophenol blue in 0·25 -Tris–HCl pH 6·8)
containing 2% 2-mercaptoethanol, before SDS–
polyacrylamide gel electrophoresis analysis.
Determination of â-galactosidase activity
â-Galactosidase activity in whole yeast cells
(measured as Miller units) was determined as
described by Zhang et al. (1991). Enzyme activity
in crude cell extracts from glass beads-broken cells
was assayed according to Guarente (1983).
Gel mobility assays
Cell extracts for gel mobility assays were prepared as described in Nasmyth and Dirick (1991).
The DNA probe was prepared by PCR amplification of a DNA fragment from base "359 to
"173 of the FRE1 promoter (Dancis et al., 1992)
using the oligonucleotides 5*-GCAGGAATTCCCAAGAACACTAAC and 5*-CTTTGAATTCCTTGAGAGAACGAT, which introduce
one EcoRI site at each end of the fragment. After
purification with a PCR Purification Kit column
(Diagen), the amplified fragment was EcoRIdigested, run in an agarose gel and extracted using
a QIAquick Gel Extraction Kit. The EcoRItreated purified fragment (100 ng) was labeled by
incubating for 15 min at 30)C in 50 ìl of Klenow
buffer (5 m-Tris–HCl pH 7·2, 10 m-MgSO4,
0·1 m-dithiothreitol) containing 1 m of each of
dCTP, dGTP and dTTP, 5 units of Klenow fragment and 70 ìCi of [32P]dATP (ICN Pharmaceuticals). Binding assays were performed in a total
volume of 20 ìl in 20 m-Tris–HCl buffer (pH 7·5)
containing 4 m- mgCl2, 50 m-NaCl, 1 mdithiothreitol, 1 ìg of bovine serum albumin,
5 m-spermidine, 5% glycerol, 2 ìg of poly(dI-dC)
and 1% Triton X-100. Twenty micrograms of total
protein from yeast extracts were diluted in this
volume of buffer and after adding one nanogram
of labeled DNA probe, the samples were incubated
at room temperature for 15 min and loaded onto
5% polyacrylamide gels (30:1 crosslinking). In
competition experiments, cold DNA probe was
added to a 50-fold molar excess.
RESULTS
Construction and characterization of aft1-deletion
strains
We have previously described RCS1 as a gene
involved in cell size control in S. cerevisiae (Gil

. 13: 621–637 (1997)
.   .
626
et al., 1991). When the published sequence was
reanalysed, it was found that the initial clone
(pMV2) from which the sequence was obtained did
not contain the 5* end of the gene nor the promoter
region. A 1·4-kb BamHI-HindIII fragment internal
to RCS1 was used as probe to isolate a new clone
(pCM4) that contained a large insert (about 10 kb)
spanning the whole RCS1 open reading frame plus
the flanking regions, from a yeast genomic DNA
library in YEp13. Plasmid pCM4 was used to
resequence the whole RCS1 open reading frame.
The open reading frame spans 2070 nucleotides,
corresponding to a product of 690 amino acids and
a calculated molecular weight of 77·4 kDa. Homology searching reveals that RCS1 is identical to
AFT1, characterized as a regulator of genes such as
FRE1, FRE2, FET3 and FTR1 (which are involved
in the high-affinity iron transport system), as well
as CCC2 and FTH1 (Yamaguchi-Iwai et al., 1995,
1996; Stearman et al., 1996). Some differences
observed between RCS1 and AFT1 sequences may
be due to strain heterogeneity.
Polyclonal antibodies raised against the last
277 amino acids of the Aft1 polypeptide were
able to detect a product of about 98 kDa in cell
extracts from a AFT1 wild-type strain, although
not from a strain carrying a aft1-Ä5 deletion (see
below; Figure 1A). A band of about 55 kDa
(probably a degradation product of Aft1)
was consistently detected in wild-type strains in
parallel to the larger band. The antibodies also
immunoprecipitated a band of 98 kDa from a
coupled transcription-translation in vitro assay
with the AFT1 DNA as template (Figure 1A),
which confirmed that the former band corresponds to Aft1. The discrepancy between the
Aft1 calculated size and that determined from its
electrophoretic mobility may be attributed to the
highly basic character of Aft1.
The aft1-Ä1 deletion (Gil et al., 1991) produces a
protein truncated at amino acid 440 (Figure 1B).
In contrast to this and other partial deletions (data
not shown), we were unable to directly construct a
haploid strain carrying a aft1-Ä5 deletion that
abolishes synthesis of Aft1 (Figure 1B). However,
this deletion was easily introduced in heterozygosis
in wild-type a/á diploid cells and from these,
haploid Ura + derivatives carrying the aft1-Ä5
allele were efficiently obtained. These aft1-Ä5 haploid cells were thus viable, although they showed a
slightly lower growth rate than isogenic wild-type
cells during the exponential phase in YPD (not
shown) or SD media (Figure 2A). Consistently,
? 1997 by John Wiley & Sons, Ltd
Figure 1. Western blots (using Aft1-directed polyclonal antibodies) of AFT1 and aft1-Ä5 cells (left) alone or transformed
with pCM20 (a plasmid that contains the EcoRI-BglII fragment
spanning the AFT1 open reading frame subcloned in
YEplac181), and autoradiography of 35S-labeled Aft1 from a
coupled transcription-translation assay with AFT1 DNA as
template before (T) and after immunoprecipitation (IP) with
the anti-Aft1 antibodies (right). Numbers at the left indicate the
mobility of standard proteins of known size (in kDa). The ca.
50 kDa specific band was not detected when extracts were
prepared as described in Printen and Sprague (1994). (B)
Restriction map of the AFT1 gene, indicating the open reading
frame of the wild-type gene and the aft1 alleles (as URA3
disruptions) employed in this work (black arrows). The fragment subcloned in pCM20 is indicated.
postdiauxic cultures of the wild-type strain reached
higher densities than those of the mutant strain.
Mutants lacking the AFT1 product are unable to
grow in media with respirable carbon sources
When testing the ability of carbon sources different from glucose to support the growth of two
strains isogenic except for the aft1-Ä5 character,
the mutant was unable to grow on respirable
carbon sources such as ethanol, glycerol or pyruvate. In contrast, it grew on fermentable carbon
sources such as glucose or fructose. Tetrad analyses indicated that the inability to grow on ethanol
plus glycerol-based medium was linked to the
aft1-Ä5 character. Glucose-grown aft1-Ä5 cells
shifted to SD-ethanol plus glycerol medium

. 13: 621–637 (1997)
      
627
deficiency. The growth defect was suppressed by
addition of ferrous ions to the medium (Figure
2B), thus showing that the inability to utilize
respirable carbon sources was caused by the poor
iron assimilation rate occurring in aft1 null mutants. At the highest concentrations tested, iron
caused a toxic effect on growth, as has been
reported previously (Yamaguchi-Iwai et al., 1995).
Mutants in the SNF1 protein kinase are unable
to derepress glucose-repressible genes, rendering
the cells unable to grow on respirable carbon
sources such as ethanol or glycerol (Carlson et al.,
1981; Celenza and Carlson, 1986; ThompsonJaeger et al., 1991). Among the SNF1-mediated
glucose-repressible genes are ADH2 (Russell et al.,
1983) and ICL1 (Fernández et al., 1992). No
defects in the induction of ADH2 or ICL1 expression upon shift from glucose to ethanol/glycerolbased medium were observed in the aft1-Ä5 cells
(data not shown). This confirmed that the carbon
source-associated growth defects of the mutant
are independent of the SNF1-mediated glucose
repression pathway.
Figure 2. Shift of wild-type AFT1 cells (CML128) and mutant
aft1-Ä5 cells (CML126) from SD-glucose medium to SDethanol plus glycerol medium. (A) CML128 (-,4) or CML126
cells (/,5) growing exponentially in SD-glucose liquid medium were centrifuged, washed and resuspended in the same
medium (-,/) or in SD-ethanol plus glycerol medium (4,5)
(time 0). Optical density at 600 nm was measured at different
times and values are plotted relative to the unit value at time 0.
(B) Effect of iron addition on growth after a shift from
SD-glucose to SD-ethanol plus glycerol. Wild-type cells (white
bars) and aft1-Ä5 cells (black bars) exponentially growing in
SD-glucose were resuspended (at an initial concentration of 106
cells per ml) in SD-ethanol plus glycerol medium added with
different concentrations of iron (as ferrous sulfate) plus ferrozine. Cell number was measured after 48 h of incubation in
these conditions. Bars indicate the percentage of the cell
concentration reached in each condition relative to the concentration reached by wild-type cells without iron or ferrozine
added. Results are from a representative experiment.
arrested growth after one to two divisions
(Figure 2A). Microscopic analysis demonstrated
that cells arrested in an unbudded state.
Since mutants in AFT1 have been shown to be
defective in iron assimilation (Yamaguchi-Iwai
et al., 1995), we determined whether the inability
of aft1-Ä5 cells to grow on ethanol plus glycerol as
the only carbon sources was exclusively due to iron
? 1997 by John Wiley & Sons, Ltd
Aft1 is differentially involved in expression of the
high-affinity iron transport system genes
Other authors have shown that a mutant in
which the AFT1 open reading frame is interrupted at codon 575 is deficient in induction of
the high-affinity iron transport genes upon iron
deprivation of the metal (Yamaguchi-Iwai et al.,
1995, 1996). However, the possibility exists that
the latter mutant still expresses a truncated Aft1
product with altered (but not absent) activity.
Thus, we studied expression of FRE1, FRE2 and
FET3 in the aft1-Ä5 null mutant. When wild-type
cells growing on glucose were deprived of iron by
addition of ferrozine at 2 m concentration,
FRE1 and FET3 expression was immediately induced, while expression of FRE2 was delayed
with respect to the other two genes (Figure 3A).
The different kinetics of expression of FRE1 and
FRE2 is in accordance with the results from
Georgatsou and Alexandraki (1994) showing that
in iron deprivation conditions FRE1-dependent
ferric reductase activity appears earlier than
FRE2-dependent activity. Iron deprivation also
induced expression of AOB629 (Figure 3A), an
open reading frame (also named YOL152w)
whose sequence has recently been determined
in the course of the project for sequencing S.
cerevisiae chromosome XV and shows significant

. 13: 621–637 (1997)
.   .
628
Figure 3. Northern blot analysis of transcripts from high-affinity
iron transport genes. (A) Wild-type AFT1 and mutant aft1-Ä5 cells
growing in SD-glucose medium were iron-deprived by addition of
2 m-ferrozine at time 0, and samples were taken for Northern
analysis at the indicated times. (B) Wild-type AFT1 cells growing in
SD-glucose medium were shifted to SD-ethanol plus glycerol medium
(time 0), and at the same time part of the cultures were iron-deprived
by addition of 2 m-ferrozine. At the indicated times, samples were
taken for Northern analysis. Equivalent amounts of RNA (as determined from ribosomal RNA staining) were run for each sample.
homology with FRE1 and FRE2 (Casamayor et
al., 1995). Levels of FRE2 mRNA were undetectable in the aft1-Ä5 mutant either in the absence
or presence of iron (Figure 3A), in accordance
with the results reported by Yamaguchi-Iwai
et al. (1995). However, the null mutant still induced FRE1 expression upon iron deprivation to
the same levels as wild-type cells, showing that
FRE1 expression is not necessarily dependent on
Aft1. AOB628 expression is also non-Aft1 dependent (Figure 3A). The aft1-Ä5 mutation affected
overall levels of expression of FET3, but not its
inducibility by iron deprivation (Figure 3A).
As for the low-affinity system for iron assimilation, FET4 mRNA levels were constitutive in
SD-glucose-grown cells, and expression of this
gene was AFT1-independent (Figure 3A).
Since we have observed that FRE1 expression
does not necessarily require Aft1, and other
authors (Yamaguchi-Iwai et al., 1996) have shown
that binding of Aft1 to the minimum consensus
sequence of the FRE1 promoter is comparatively
? 1997 by John Wiley & Sons, Ltd
less intense than to other high-affinity transport
genes, gel shift assays were carried out with a
probe comprising a larger region of the FRE1
promoter previously shown to contain ironresponsive elements (Dancis et al., 1992). Cell
extracts from wild-type and mutant cultures
grown in the presence and absence of iron were
assayed (Figure 4). Two specific retardation complexes were observed in cells grown with iron,
that were Aft1-independent. These same complexes were formed in wild type cells in conditions in which the FRE1 promoter is induced
by iron deprivation. In contrast, extracts from
iron-deprived mutant cells did not form the indicated retardation complexes, but instead other
specific ones of higher mobility could be detected in these conditions. Electrophoretic analysis
demonstrated that protein degradation had not
occurred in any of the extracts (not shown). We
conclude that the larger promoter region of
FRE1 is able to bind specific protein complexes
in an Aft1-dispensable way.

. 13: 621–637 (1997)
      
629
initial lag (see Figure 1). Although they were more
sensitive to iron deprivation than wild-type cells,
aft1-Ä5 cells were also unable to derepress FRE1
or FRE2 expression merely by shifting glucosegrown cells to ethanol/glycerol (not shown).
Ferrozine-treated mutant cells induced FRE1 expression after 24 h of being shifted to ethanol/
glycerol (not shown).
The low-affinity iron transport system probably
does not play a significant role in ethanol/glycerolgrown cells, since expression of FET4 decreases in
these conditions to almost undetectable levels with
respect to glucose-grown cells (Figure 3B).
Figure 4. Gel retardation assays with extracts from wild-type
AFT1 and mutant aft1–5 cells and a labelled 187 bp fragment
from the FRE1 promoter as probe. Cells were exponentially
grown in SD-glucose medium, and part of the cultures were
incubated for 8 h in the presence of 1 m-ferrozine ("Fe)
while the other part remained intact for the same time interval
(+Fe), before extract preparation. For each extract, two assays
were carried out in parallel, one with labelled probe alone and
another with a 50-fold molar excess of unlabelled probe also
added to the mixture. The reaction contents were run in
polyacrylamide gels and these were autoradiographed. The
left-most lane corresponds to labelled probe alone. The arrows
point to the two more intense specific retardation bands.
Shift from glucose to ethanol plus glycerol does
not induce expression of genes involved in iron
assimilation
Our results and those of other authors (Askwith
et al., 1994) show that yeast cells altered in iron
assimilation are unable to carry out respiratory
metabolism adequately. Thus, since yeast cells
growing on respirable carbon sources require
higher cytoplasmic iron levels than on fermentable
carbon sources, the possibility exists that shifting
cells from glucose to ethanol plus glycerol-based
medium will induce expression of iron
assimilation-related genes such as FRE1, FRE2 or
FET3.
Shift of wild-type cells from SD-glucose to SDethanol plus glycerol without changing iron availability did not induce FRE1, FRE2, AOB629 or
FET3 expression even after 48 h in the latter
conditions (Figure 3B). When cells were irondeprived at the time of the nutritional shift, induction of FRE1, FRE2 or FET3 occurred, although
only after 24 h of ferrozine addition (Figure 3B)
concomitantly with resumption of growth after an
? 1997 by John Wiley & Sons, Ltd
Size increase in cells with a truncated AFT1
product is iron-suppressible
Mutant cells carrying the aft1-Ä1 allele (that
produces an Aft1 polypeptide lacking the
C-terminal 250 amino acids) are larger than wildtype cells when growing on glucose (Gil et al.,
1991), a phenotype that is not observed in the
aft1-Ä5 null mutant. The former mutant cells are
as sensitive to iron deprivation as aft1-Ä5 cells.
Thus, while growth of wild-type cells (OL1 and
CML128 strains) in SD-glucose medium is only
affected at 0·5 m or higher ferrozine concentrations, aft1-Ä1 cells (strain CML12) or aft1-Ä5
cells (strain CML126) are sensitive to ferrozine
concentrations from 0·05 m. We tested if the cell
size alteration of the aft1-Ä1 cells is suppressed by
addition of iron to the medium. In fact, addition
of 0·4 m-ferrous sulfate plus 1 m-ferrozine
(conditions where wild-type and aft1-Ä1 cells
grow with a doubling time of 120 min, the same
as without iron/ferrozine addition) restored an
aft1-Ä1 mutant average cell size equivalent to
that of wild-type cells (Table 2), indicating that
the cell size defect in the mutant is caused by iron
deficiency. This mutant behaves similarly to the
null aft1-Ä5 mutant with respect to expression of
FRE1, FRE2 or FET3 in the presence or absence
of iron (not shown).
Properties of Aft1 as a transactivator
Since Aft1 regulates the expression of the highaffinity iron transport genes through its binding
capability to consensus promoter sequences, we
studied the possibility that the Aft1 property as a
transcriptional activator could be directly modulated by iron levels. For this purpose, we made a
construction coding for Gal4bd fused in frame to
the Aft1 polypeptide (see Materials and Methods)

. 13: 621–637 (1997)
.   .
630
Table 2.
Effect of iron addition on the size of aft1-Ä1 cells.
Strain
OL1
OL1
CML12
CML12
Fe/ferrozine (m) added
Average cell volume (ìm3)&SD
None
0·4/1·0
None
0·4/1·0
53·1&20·9
65·9&15·9
82·1&21·5
65·6&18·7
Cells were grown exponentially for at least ten generations in SD-glucose medium without or with
ferrous sulfate (Fe) plus ferrozine added, before fixation with 1% formaldehyde and microscopic
observation.
and introduced it in plasmid pCM86. Y153 cells
(containing a GAL1-HIS3 reporter system whose
expression is driven by Gal4p DNA binding plus
activating sequences; Durfee et al., 1993) transformed with pCM86 were able to grow on SDglucose plates lacking histidine and containing
3-aminotriazole, in contrast with Y153 cells transformed with the control plasmid pGBT9 (not
shown). This confirmed that Aft1 also has activating properties when fused to a different binding
domain. Moderated activating capability of Aft1
was also detected when transcription of the GAL1lacZ reporter system in SD-glucose-grown cells
was measured (Table 3). Iron concentration did
not have any effect on the transcriptional activity
of Gal4bd-Aft1 on the GAL1-lacZ reporter system. In fact, â-galactosidase activity was not
affected in cells growing on glucose after 8 h in
the presence of ferrozine (Table 3).
When cells were shifted from SD-glucose to
SD-ethanol plus glycerol medium, Aft1 displayed
an enhanced transcriptional activity, equivalent to
that of a fusion of Gal4bd to the viral VP16
transcriptional activation domain (ad) in the same
growth conditions (Table 3). The Gal4bd-VP16ad
fusion also showed higher activity in ethanol plus
glycerol-grown cells than in glucose-grown cells,
confirming that the GAL1-lacZ system present in
pGBT9 is subjected to glucose repression and
therefore, that glucose-grown cells are not the
optimal conditions to measure transcriptional activation in this system. Glucose repression could be
acting through the URS sequences still present
in the GAL1 promoter construction (Flick and
Johnston, 1990; Johnston et al., 1994). As in
glucose-grown cells (see above), iron deprivation
did not have any effect on the transcriptional
activity of Gal4bd-Aft1 in ethanol-plus glycerolgrown cells (not shown).
Aft1 is phosphorylated depending on the
nutritional state of the cells
We next studied the possible correlations between the properties of Aft1 as a transcriptional
Table 3. In vivo transcriptional activity of Aft1 on a GAL1-lacZ reporter system in
different growth conditions.
â-Galactosidase activity in cells transformed witha
Growth conditions
SD-glucose medium
SD-glucose medium plus
1 m-ferrozineb
SD-ethanol/glycerol mediumc
pGBT9
pCM86
pCM110
<0·1
1·2
7·9
<0·1
0·4
0·9
76·9
11·2
76·2
a
Exponentially growing SFY526 cells transformed with pGBT9 (control), pCM86 (Gal4bd-Aft1) or
pCM110 (Gal4bd-VP16ad) were analysed for enzyme activity (Miller units) after permeabilization.
Results are from a representative experiment.
b
After 8 h in the presence of the chelator.
c
After 72 h of shift from SD-glucose medium.
? 1997 by John Wiley & Sons, Ltd

. 13: 621–637 (1997)
      
631
Figure 5. Western blot analysis of Aft1 in different growth conditions. Lower and upper arrows correspond to
the under- and hyperphosphorylated forms of Aft1 respectively. (A) Wild-type cells in SD-glucose medium added
with 1 m-ferrozine at time 0. After 4 h in these conditions, cells were diluted one fourth in fresh medium with
ferrozine to keep them exponentially dividing. (B) Wild-type cells growing exponentially in SD-glucose and shifted
at time 0 to SD-ethanol plus glycerol. (C) Wild-type cells growing exponentially in YEPD medium and shifted at
time 0 to YEP medium. (D). Immunoprecipitated Aft1 from wild-type cells growing exponentially in YEPD
medium treated with calf intestine alkaline phosphatase (CIP) and a mixture of inhibitors of the enzyme. (E)
Wild-type AFT1 and mutant aft1-Ä5 cells growing exponentially in YEPD medium (e) or after having traversed
the diauxic shift (d).
factor and changes in the protein depending on the
growth conditions. To determine whether Aft1 is
post-transcriptionally modified when cells are deprived of iron, cells were grown in SD-glucose
medium plus 1 m-ferrozine, which are conditions
that induce FRE1, FRE2 or FET3 expression (see
above) but do not affect cell division kinetics
during the next 10 h after iron deprivation (higher
concentrations of ferrozine affect cell growth after
8 h, data not shown in detail). Aft1 was not
modified in iron-deprived cells that were dividing
normally (Figure 5A), at least in a way that could
affect its electrophoretic mobility. In contrast, in
cells shifted from SD-glucose to SD-ethanol plus
glycerol medium, a band of lower mobility (ca.
100 kDa) than the 98 kDa band was also recognized by the anti-Aft1 antibodies (Figure 5B). This
band appeared during the initial 24 h in the ethanol plus glycerol medium, when cells were not
growing, although not in samples from later times
when cells had resumed growth. The 100 kDa form
was also formed when cells growing in YPD
medium were shifted to YP medium deprived of
carbon source (Figure 5C), and it must correspond
to Aft1 since it did not appear in extracts from
aft1-Ä5 cells (Figure 5E). Alkaline phosphatase
treatment caused disappearance of the lower mobility band, a fact that did not occur in the
presence of inhibitors of the enzyme (Figure 5D).
This demonstrates that the 100 kDa band corresponds to a hyperphosphorylation state of Aft1.
? 1997 by John Wiley & Sons, Ltd
Nutrient deprivation at the end of the exponential phase in the yeast population growth cycle
causes a diauxic shift, that is, a change from
fermentative to respiratory metabolism (WernerWashburne et al., 1993). At the time of the diauxic
shift glucose concentration remains considerably
high (François et al., 1987; and our unpublished
results), so that deprivation of other nutrients must
be responsible for induction of the shift. We observed that Aft1 is also phosphorylated when cells
traverse the diauxic shift stage at the end of the
exponential phase, and remains hyperphosphorylated until stationary phase (Figure 5E).
Since Aft1 is phosphorylated when cells become
deprived of glucose, we tested if this modification
is dependent on the SNF1 protein kinase. This is
not the case, since a mutant lacking SNF1 was still
able to phosphorylate Aft1 during nutritional
shifts (data not shown). It might be conceivable
that phosphorylation of Aft1 upon glucose deprivation could be antagonized by maintenance of
constitutive high levels of cAMP-mediated protein
kinase A activity (Toda et al., 1987), which normally decreases drastically after glucose exhaustion (Werner-Washburne et al., 1993; Thevelein,
1994). However, a cyr1 mutant (strain TC41–1)
externally added with 1 m-cAMP still phosphorylated Aft1 (data not shown in detail), a result
that discards the above hypothesis.
Phosphorylation of Aft1 is not only induced by
changes in the nutritional state of the cells. In fact,

. 13: 621–637 (1997)
.   .
632
Figure 6. Effect of iron on cells overexpressing AFT1. (Left panel) CML72 cells transformed with
YCpGAL or with pCM49 (YCpGAL-AFT1) were grown exponentially in SD liquid medium plus
raffinose and then resuspended (at a concentration of 106 cells per ml) in SD liquid medium plus
raffinose alone (SD-Raf) or with raffinose plus galactose (SD-Raf+Gal), added with the indicated
amounts of BPS. Cell number was measured after 24 h at 30)C, and values are given relative to the
cell concentration reached by YCpGAL transformants in SD-Raf medium. (Right panel) As above
except that iron (as ferrous sulfate) and ferrozine were added at the indicated concentrations.
the Aft1 protein also became phosphorylated in
cdc28 and cdc25 cells at the non-permissive temperature, conditions in which cells arrest growth at
the G1 stage of the cell cycle. In contrast, other G1
arrest conditions such as á-factor treatment (that
does not affect growth) did not result in phosphorylation of Aft1 (data not shown in detail).
Overexpression of AFT1 arrests cell growth at G1
To study further the role of Aft1 we made a
construction in which AFT1 was put under the
control of the GAL1 promoter in a centromeric
plasmid. Galactose-induced overexpression of
Aft1 caused inhibition of cell growth in the host
strain (Figure 6). Flow cytometry analysis showed
that after 24 h of induction conditions, cells had
arrested with a haploid DNA content, and microscopic observations indicated that in these conditions cells remained in an unbudded state (not
shown in detail). That is, proliferation arrest in
galactose-grown cells occurs at the G1 stage of the
cell cycle.
Growth arrest in AFT1-overexpressing cells
could be caused by the toxic effect of an excess of
intracellular iron and/or copper ions as a consequence of unregulation of the Aft1-mediated highaffinity iron transport system (which is also
involved in copper entrance; Hassett and Kosman,
1995; Askwith et al., 1996). If this were the case,
reduction of the levels of available iron and copper
in the medium could suppress the toxicity of AFT1
overexpression. Such a possibility must be dis? 1997 by John Wiley & Sons, Ltd
carded since adding BPS (a chelator of iron and
copper ions; Stearman et al., 1996) to the medium
did not suppress Aft1 overexpression-caused
growth arrest (Figure 6, left).
On the other hand, AFT1 overexpression could
result in increased expression of some protein
leading to growth arrest. In case iron could directly
inhibit Aft1 activity, addition of extra iron to the
medium would be expected to suppress growth
arrest. However, iron addition (as well as that of
copper, not shown) did not allow growth of cells
overproducing Aft1 (Figure 6, right). On the
contrary, externally added iron enhanced the
growth-inhibitory effect of AFT1 overexpression
at concentrations that still were not toxic for
cells with normal levels of AFT1 expression.
These results also discard the possibility that Aft1
titrates some component participating in an
Aft1-independent iron uptake mechanism whose
functionality otherwise would be required for
growth in the medium conditions employed for
overexpression.
DISCUSSION
The product of AFT1 is a transcriptional factor
that regulates expression of the FRE1 and FRE2
ferric reductases and the components of the transport complex responsible for the high-affinity iron
uptake in S. cerevisiae (Yamaguchi-Iwai et al.,
1995, 1996; Stearman et al., 1996). Previously, we
have shown that a mutant lacking the 3*-terminal

. 13: 621–637 (1997)
      
part of the gene has an average cell size larger than
normal (Gil et al., 1991). In this work, we demonstrate that this mutant is also deficient for iron
uptake and that the size defect is suppressed by the
addition of iron to the medium, therefore correlating both phenotypes. These observations indicate
that a deficiency in iron may slow passage of cells
through the G1 stage of the cell cycle, leading to a
larger cell size at Start, and therefore to an increase
of the average cell size in the population. Complete
deprivation of iron causes homogeneous cell
growth arrest at G1 (our unpublished observations), as occurs with deprivation of nutrients such
as nitrogen, phosphorous or sulphur (WernerWashburne et al., 1993). It is remarkable that a
null AFT1 mutant also deficient in intracellular
iron accumulation has not the same size defect as
the mutant producing the truncated protein, thus
pointing to a more complex correlation between
the Aft1 role in iron uptake and cell cycle progression. In any case, the partially dominant phenotype of the mutant producing the truncated Aft1
protein (Gil et al., 1991) may be an indication of
interactions of Aft1 with components of multiprotein complexes that would be functionally inactivated by the truncated product, a situation not
occurring in cells lacking Aft1. In this latter case,
cells would be prone to cell cycle advance (maybe
because other unknown elements substitute the
function of Aft1) but not to efficient high-affinity
iron uptake.
The inability of AFT1 loss-of-function mutants
to grow on respiratory carbon sources can be
explained by the requirement of higher amounts of
intracellular iron to support respiratory growth
compared with fermentative growth. Basal levels
of expression of FET3 in the aft1-Ä5 background
(this work) are concomitant with some transporter
activity remaining in the absence of Aft1
(Yamaguchi-Iwai et al., 1995), and they could
be sufficient to support fermentative growth although not respiratory growth. Two additional
points must be made. First, a double Äfre1 Äfre2
mutant still shows some ferric reductase activity
(Georgatsou and Alexandraki, 1994), pointing to
the existence of other genes coding for enzymes
able to reduce Fe(III) to Fe(II). The product of the
open reading frame AOB629 could be one of them,
and since its expression is not AFT1-dependent, it
could provide enough Fe(II) ions to aft1-Ä5 cells.
Second, the FET4-mediated low-affinity system for
ion transport, which mainly utilizes Fe(II) as substrate (Dix et al., 1994), is operative in the absence
? 1997 by John Wiley & Sons, Ltd
633
of Aft1 (this work) and therefore could provide
some intracellular iron at the concentrations of the
metal present in the SD medium. On ethanolglycerol medium, FET4 expression is lowered and
consequently, contribution of the low-affinity system to intracellular iron accumulation could be
greatly diminished. In any case, the high-affinity
iron transport system seems to be the essential one
operating both in respiratory and fermentative
conditions, which explains that even in SD-glucose
medium Aft1-lacking cells grow slower than
wild-type ones. Besides affecting such diverse
physiological processes as cytochrome or deoxyribonucleotide biosynthesis (Wrigglesworth and
Baum, 1980), low intracellular iron levels could
also alter glycolysis flux through the posttranscriptional positive regulatory role of this
metal on TPI3-coded triosephosphate isomerase
and THD3-coded glyceraldehyde-3-phosphate isomerase (Krieger and Ernst, 1994).
Mutant aft1-Ä5 cells are hypersensitive to iron
deprivation while still keeping the ability to induce
FRE1 or FET3 expression upon ferrozine addition.
In SD medium containing iron, they do not show
higher levels of expression of these genes than
wild-type cells in conditions (respiratory carbon
sources) where intracellular iron levels are so low
as to be unable to support cell growth. The result
indicates that low internal levels of iron alone
cannot act as a signal for induction of the highaffinity transport genes, and suggests that this
induction requires a sensor [either the permease/
transport complex (Stearman et al., 1996) or some
uncharacterized component] able to measure external iron concentrations in order to transfer the
signal to the transcriptional complex where Aft1
participates.
Yamaguchi-Iwai et al. (1995, 1996) have demonstrated the role of Aft1 in mediating expression of
the high-affinity iron uptake genes through its
ability to bind to a specific sequence of the promoter of these genes. Our results confirm that Aft1
is a transcriptional factor able to activate in vivo
the expression of two reporter systems when fused
to a different DNA binding moeity. Two points
deserve comment when relating these results to the
role of Aft1 on expression of the iron uptake
genes. First, iron starvation does not increase
Aft1-mediated expression of the lacZ reporter system, indicating that the transcriptional activation
capability of Aft1 is iron-independent, and suggesting that only binding of the transcriptional
complex to the Aft1-binding promoter sequences

. 13: 621–637 (1997)
.   .
634
(Yamaguchi-Iwai et al., 1996) is iron-modulated
through other compoments of the complex.
Second, since the aft1-Ä1 mutant is still irondeficient, the C-terminal part of the Aft1 protein
must be essential for expression of the iron transport system. This C-terminal region is especially
rich in glutamine residues, which might be required
for the transactivation properties of Aft1
(Yamaguchi-Iwai et al., 1995).
Studies with the aft1-Ä5 mutant demonstrate
that Aft1 is differentially involved in regulating the
expression of the high-affinity iron uptake genes.
Thus, Aft1 is essential for providing basal (plus
iron) and inducible (minus iron) levels of FRE2
expression, as well as basal levels of FET3 expression. However, this latter gene is still inducible in
an aft1-Ä5 background in the absence of iron.
With respect to FRE1, no differences in expression were observed between the wild-type strain
and the null mutant both in iron-plus or irondepleted conditions, indicating that the Aft1 transcriptional factor is not essential for its expression.
This is in contrast with the results from other
authors (Yamaguchi-Iwai et al., 1995) using a
different aft1 mutant. We propose that the mutant
employed in the latter work is not a total loss-offunction one but one that expresses an Aft1 protein truncated at amino acid 575, that could
interfere with Aft1-regulated gene expression in a
dominant way. Our truncated aft1-Ä1 mutant also
exhibits partially dominant effects (see above). The
observations in the present work do not contradict
the fact that in normal conditions Aft1 would
participate in FRE1 transcriptional complexes, as
shown by the phenotype of a constitutively upregulated AFT1 mutant (Yamaguchi-Iwai et al.,
1995). However, in the absence of Aft1 some other
transcriptional factor could substitute it efficiently.
Remarkably, in gel mobility shift assays Aft1
shows comparatively much lower binding ability
to the FRE1 promoter Aft1-binding site than to
the sites of other high-affinity iron uptake genes
(Yamaguchi-Iwai et al., 1996).
By using the whole iron-responsive element of
the FRE1 promoter (which includes the Aft1binding site; Dancis et al., 1992), we have observed
that: (i) in the presence of iron there are no
differences in the mobility of the promoter complexes between mutant and wild-type cells, and (ii)
higher-mobility complexes are observed in the mutant in the absence of iron. Thus, Aft1 would only
be required to form transcriptional complexes in
the absence of iron. However, as the aft1 null
? 1997 by John Wiley & Sons, Ltd
mutant is perfectly able to induce FRE1 expression
upon iron deprivation, Aft1-independent highermobility complexes in iron-deprived cells would
still be transcriptionally active, pointing to the
presence of other transcriptional factors on the
FRE1 promoter.
The product of MAC1 is a transcriptional factor
involved in copper uptake that is also required for
basal expression of FRE1, although not for its
inducibility upon iron deprivation (Jungmann et
al., 1993; Hassett and Kosman, 1995), a situation
that is reminiscent of that of AFT1 with respect to
FET3. The possibility is raised that Mac1 could
substitute Aft1 for FRE1 transcription only in the
absence of the latter, although other phenotypes of
the respective null mutants differentiate both factors. Thus, a MAC1 null mutant is comparable to
aft1-Ä5 cells in being unable to grow on respiratory
carbon sources (Jungmann et al., 1993), but this
situation is non-iron-suppressible in contrast to the
AFT1 mutant. Also, while cells lacking Mac1 are
hypersensitive to oxidative stress (Jungmann et al.,
1993), this is not the case for aft1-Ä5 mutants (our
unpublished results). Therefore, in normal conditions MAC1 and AFT1 do not seem to participate
in the same pathway in relation to iron uptake.
The phenotype of the AFT1 null mutant with
respect to expression of genes for high-affinity iron
transport points to a situation where different
factors may be involved in transcriptional complexes for the different genes. Other studies have
shown that although having partially resembling
elements (among them the Aft1-binding element;
Yamaguchi-Iwai et al., 1996), FRE1 and FRE2
promoters are differentially dependent on the
YAP1 transcriptional factor (Georgatsou et al.,
1995; Lesuisse and Labbe, 1995). Yap1 is also
involved in oxidative stress responses (Stephen et
al., 1995). Thus, transcriptional factors such as
Yap1, Mac1 and Aft1 may be shared by some but
not all the complexes interacting with the promoters of the iron-uptake genes, at the same time that
these factors may also participate in other cellular
processes such as stress responses. FRE1 ferric
reductase activity is induced earlier than FRE2
activity upon iron deprivation (Georgatsou and
Alexandraki, 1994), in accordance with our results
on synthesis of the respective mRNAs. The timelydifferentiated induction of the iron-uptake genes
also supports the idea of separate complexes
inducing the respective genes.
Overexpression of AFT1 leads to uniform cell
growth arrest at G1. This effect is not caused by

. 13: 621–637 (1997)
      
iron or copper toxicity due to increased uptake of
these metals, nor by inhibition of iron transport in
cells overproducing Aft1, thus pointing to the
participation of the latter in other cellular processes besides iron assimilation. As Aft1-mediated
growth inhibition is not counteracted by iron
addition, the possibility that iron is a direct inhibitor of Aft1 activity can be discarded. Conditions
that temporarily or permanently arrest cell growth
(for instance, carbon source deprivation, shift
from fermentative to respiratory metabolism, cell
cycle arrest at G1 in cdc mutants) cause Aft1
phosphorylation by a mechanism that is independent of the SNF1 protein kinase as well as from
protein kinase A. Phosphorylation also occurs
when cells change from fermentative to respiratory
metabolism at the diauxic shift of the population
growth cycle. On the other hand, transitory growth
arrest at G1 by alpha factor treatment (which does
not cause a marked growth arrest) does not lead to
Aft1 phosphorylation. Therefore, correlation between Aft1 phosphorylation and growth arrest
seems likely to exist.
During a shift to ethanol-glycerol medium
(where a transitory growth arrest occurs and Aft1
becomes
temporarily
hyperphosphorylated),
FRE1, FRE2 and FET3 expression is induced upon
iron deprivation only when growth has resumed
and Aft1 is again in the underphosphorylated
state. This suggests that phosphorylation may be a
mechanism for Aft1 inactivation. By forming part
of the respective transcriptional complexes, Aft1
may be a mediator of iron uptake induction (as
well as other putative cellular processes) only in
conditions where cells are able to grow once metabolic adaptation to the available carbon source has
occurred. The adaptation to the available carbon
source would be metabolically dominant over
other nutrient signals, Aft1 playing a role in linking both signals.
ACKNOWLEDGEMENTS
We thank D. Alexandraki, G. Berben, A. Boyd, L.
del Castillo, F. Estruch and J. Zueco for the
generous gift of plasmids and strains, N. Colomina
and F. Purroy for collaborating in some experiments, X. Gómez for introducing us to cytofluorimetric techniques and E. Garí for critically reading
the manuscript. The technical skill of L. Piedrafita
is especially appreciated. C. C. was the recipient of
grants from the European Union and the Ajuntament de Lleida. C. G. was partially supported by a
? 1997 by John Wiley & Sons, Ltd
635
post-doctoral grant from the Spanish Ministry of
Education and Science. This work was supported
by DGICYT (grants PB91–0237 and PB94–0511),
CIRIT (grant GRQ93–6003) and the Ajuntament
de Lleida.
REFERENCES
Askwith, C., de Silva, D. and Kaplan, J. (1996). Molecular biology of iron acquisition in Saccharomyces
cerevisiae. Mol. Microbiol. 20, 20–27.
Askwith, C., Eide, D., Van Ho, A., et al. (1994). The
FET3 gene of S. cerevisiae encodes a multicopper
oxidase required for ferrous iron uptake. Cell 76,
403–410.
Ausubel, F. M., Brent, R., Kingston, R. E., et al. (1989).
Current Protocols in Molecular Biology. WileyInterscience, New York.
Bagg, A. and Neilands, J. B. (1987). Molecular mechanisms of regulation of siderophore-mediated iron
assimilation. Microbiol. Rev. 51, 509–518.
Bartel, P. L., Chien, C. T., Sternglans, R. and Fields, S.
(1993). Using the two-hybrid system to detect proteinprotein interactions. In Hartley, D. A. (Ed.), Cellular
Interactions in Development: A Practical Approach.
Oxford University Press, Oxford, pp. 153–179.
Berben, G., Dumont, J., Gilliquer, V., Bolle, P. A. and
Hilger, F. (1991). The YDp plasmids: a uniform set of
vectors bearing versatile gene disruption cassettes for
Saccharomyces cerevisiae. Yeast 7, 475–477.
Carlson, M., Osmond, B. C. and Botstein, D. (1981).
Mutants of yeast defective in sucrose utilization.
Genetics 98, 25–40.
Casamayor, A., Aldea, M., Casas, C., et al. (1995).
DNA sequence analysis of a 13 kbp fragment of the
left arm of yeast chromosome XV containing seven
new open reading frames. Yeast 11, 1281–1288.
Celenza, J. L. and Carlson, M. (1986). A yeast gene that
is essential for release from glucose repression encodes
a protein kinase. Science 233, 1175–1180.
Crichton, R. R. (1991). Inorganic Biochemistry of Iron
Metabolism. Ellis Horwood Inc., Chichester, UK.
Dancis, A., Klausner, R. D., Hinnenbusch, A. G. and
Barriocanal, J. G. (1990). Genetic evidence that ferric
reductase is required for iron uptake in Saccharomyces cerevisiae. Mol. Cell. Biol. 10, 2294–2301.
Dancis, A., Roman, D. G., Anderson, G. J.,
Hinnebusch, A. G. and Klausner, R. D. (1992). Ferric
reductase of Saccharomyces cerevisiae: molecular
characterization, role in iron uptake, and transcriptional control by iron. Proc. Natl. Acad. Sci. USA 89,
3869–3873.
Dancis, A., Yuan, D. S., Halle, D., et al. (1994).
Molecular characterization of a copper transport protein in S. cerevisiae: an unexpected role for copper in
iron transport. Cell 76, 393–402.
De Silva, D. M., Askwith, C. C., Eide, D. and Kaplan,
J. (1995). The FET3 gene product required for high

. 13: 621–637 (1997)
.   .
636
affinity iron transport in yeast is a cell surface ferroxidase. J. Biol. Chem. 270, 1098–1101.
Dix, D. R., Bridham, J. T., Broderius, M. A.,
Byersdorfer, C. A. and Eide, D. J. (1994). The FET4
gene encodes for the low affinity Fe(II) transport
protein of Saccharomyces cerevisiae. J. Biol. Chem.
269, 26092–26099.
Durfee, T., Becherer, K., Chen, P. L., et al. (1993). The
retinoblastoma protein associates with the protein
phosphatase type 1 catalytic subunit. Genes Dev. 7,
555–569.
Eide, D., Davis-Kaplan, S., Jordan, I., Sipe, D. and
Kaplan, J. (1992). Regulation of iron uptake in Saccharomyces cerevisiae: the ferrireductase and Fe(II)
transporter are regulated independently. J. Biol.
Chem. 267, 20774–20781.
Engler-Blum, G., Meier, M., Frank, J. and Muller,
G. A. (1993). Reduction of background problems in
nonradioactive northern and southern blot analyses
enables higher sensitivity than 32P-based hybridizations. Anal. Biochem. 210, 235–244.
Fernández, E., Moreno, F. and Rodicio, R. (1992). The
ICL1 gene from Saccharomyces cerevisiae. Eur. J.
Biochem. 204, 983–990.
Flick, J. S. and Johnston, M. (1990). Two systems of
glucose repression of the GAL1 promoter in Saccharomyces cerevisiae. Mol. Cell. Biol. 10, 4757–4769.
François, J., Eraso, P. and Gancedo, C. (1987). Changes
in the concentration of cAMP, fructose 2,6bisphosphate and related metabolites and enzymes in
Saccharomyces cerevisiae during growth on glucose.
Eur. J. Biochem. 164, 369–373.
Georgatsou, E. and Alexandraki, D. (1994). Two distinctly regulated genes are required for ferric reduction, the first step of iron uptake in Saccharomyces
cerevisiae. Mol. Cell. Biol. 14, 3065–3073.
Georgatsou, E., Mavrogiannis, L., Frangiadakis, G.,
Klinakis, A. and Alexandraki, D. (1995). The Saccharomyces cerevisiae ferric reductase genes FRE1 and
FRE2 are distinctly regulated by iron, copper and
stress related factors. Yeast 11, s155.
Gietz, R. D., St. Jean, A., Woods, R. A. and Schiestl,
R. H. (1992). Improved method for high efficiency
transformation of intact yeast cells. Nucl. Acids Res.
220, 1425.
Gietz, R. D. and Sugino, A. (1988). New yeast–
Escherichia coli shuttle vectors constructed with in
vitro mutagenized yeast genes lacking six-base pair
restriction sites. Gene 74, 527–534.
Gil, R., Zueco, J., Sentandreu, R. and Herrero, E.
(1991). RCS1, a gene involved in controlling cell size
in Saccharomyces cerevisiae. Yeast 7, 1–14.
Gollwitz, D. and Suves, J. (1980). Complete nucleotide
sequence of the actin gene in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 77, 2545–2550.
Gossen, M. and Bujard, H. (1992). Tight control of gene
expression in mammalian cells by tetracyclin? 1997 by John Wiley & Sons, Ltd
responsive promoters. Proc. Natl. Acad. Sci. USA 89,
5547–5551.
Guarente, L. (1983). Yeast promoters and lacZ fusions
designed to study expression of cloned genes in yeast.
Methods Enzymol. 101, 181–191.
Guerinot, M. L. (1994). Microbial iron transport. Annu.
Rev. Microbiol. 48, 743–772.
Harshman, K. D., Moye-Rowley, W. S. and Parker,
C. S. (1988). Transcriptional activation by the SV40
AP-1 recognition element in yeast is mediated by a
factor similar to AP-1 that is distinct from GCN4.
Cell 53, 321–330.
Hassett, R. and Kosman, D. J. (1995). Evidence for
Cu(II) reduction as a component of copper uptake by
Saccharomyces cerevisiae. J. Biol. Chem. 270, 128–
134.
Huble, L., Bradshaw-Rouse, J. and Heideman, W.
(1993). Connections between the Ras-cyclic AMP
pathway and G1 cyclin expression in the budding
yeast Saccharomyces cerevisiae. Mol. Cell. Biol. 13,
6274–6282.
Johnston, M., Flick, J. S. and Pexton, T. (1994). Multiple mechanisms provide rapid and stringent glucose
repression of the GAL gene expression in Saccharomyces cerevisiae. Mol. Cell. Biol. 14, 3834–3841.
Jungmann, J., Reins, H. R., Lee, J., et al. (1993). MAC1,
a nuclear regulatory protein related to Cu-dependent
transcription factors is involved in Cu/Fe utilization
and stress resistance in yeast. EMBO J. 12, 5051–
5056.
Krieger, K. and Ernst, J. F. (1994). Iron regulation of
triosephosphate isomerase transcript stability in the
yeast Saccharomyces cerevisiae. Microbiology 140,
1079–1084.
Kusters, J. G., Jager, E. J. and van der Zeijst, B. A. M.
(1989). Improvement of the cloning linker of the
bacterial expression vector pEX. Nucl. Acids Res. 17,
8007.
Lee, K. S. and Levin, D. E. (1992). Dominant mutations
in a gene encoding a putative protein kinase (BCK1)
bypass the requirement for a Saccharomyces cerevisiae
protein kinase C homolog. Mol. Cell. Biol. 12, 172–
182.
Lesuisse, E. and Labbe, P. (1989). Reductive and nonreductive mechanisms of iron assimilation by the yeast
Saccharomyces cerevisiae. J. Gen. Microbiol. 135, 257–
263.
Lesuisse, E. and Labbe, P. (1995). Effects of cadmium
and of YAP1 and CAD1/YAP2 genes on iron
metabolism in the yeast Saccharomyces cerevisiae.
Microbiology 141, 2937–2943.
Makowski, G. S. and Ramsby, M. L. (1993). pH modification to enhance the molecular sieving properties of
sodium dodecyl sulfate-10% polyacrylamide gels.
Anal. Biochem. 212, 283–285.
Nasmyth, K. and Dirick, L. (1991). The role of SWI4
and SWI6 in the activity of G1 cyclins in yeast. Cell
66, 995–1013.

. 13: 621–637 (1997)
      
Nasmyth, K. and Tatchell, K. (1980). The structure of
transposable yeast mating type loci. Cell 19, 753–764.
Nieuwint, R. F. M., Mager, W. H., Maurer, K. C. T.
and Planta, R. J. (1989). Mutational analysis of the
upstream activation site of yeast ribosomal genes.
Curr. Genet. 15, 247–251.
Printen, J. A. and Sprague, G. F. Jr. (1994). Protein–
protein interactions in the yeast pheromone response
pathway: Ste5p interacts with all members of the
MAP kinase cascade. Genetics 138, 609–619.
Rose, M. D., Winston, F. and Hieter, P. (1990). Methods
in Yeast Genetics: a Laboratory Course Manual.
Cold Spring Harbor Laboratory Press, Cold Spring
Harbor, NY.
Rothstein, R. (1983). One-step gene disruption in yeast.
Methods Enzymol. 101, 202–211.
Russell, D. W., Smith, M., Williamson, V. M. and
Young, E. T. (1983). Nucleotide sequence of the yeast
alcohol dehydrogenase II gene. J. Biol. Chem. 258,
2674–2682.
Sambrook, J., Fritsch, E. F. and Maniatis, T. (1989).
Molecular Cloning: a Laboratory Manual, 2nd edn.
Cold Spring Harbor Laboratory Press, Cold Spring
Harbor, NY.
Sanger, F., Nicken, S. and Coulson, A. R. (1977). DNA
sequencing with chain-terminating inhibitors. Proc.
Natl. Acad. Sci. USA 74, 5463–5467.
Stearman, R., Yuan, D. S., Yamaguchi-Iwai, Y.,
Klausner, R. D. and Dancis, A. (1996). A permeaseoxidase complex involved in high-affinity iron uptake
in yeast. Science 271, 1552–1557.
Stephen, D. W. S., Rivers, S. L. and Jamieson, D. J.
(1995). The role of the YAP1 and YAP2 genes in the
regulation of the adaptative oxidative stress responses
of Saccharomyces cerevisiae. Mol. Microbiol. 16, 415–
423.
Thevelein, J. M. (1994). Signal transduction in yeast.
Yeast 10, 1753–1790.
? 1997 by John Wiley & Sons, Ltd
637
Thompson-Jaeger, S., François, J., Gaughran, J. P. and
Tatchell, K. (1991). Deletion of SNF1 affects the
nutrient response of yeast and resembles mutations
which activate the adenylate cyclase pathway. Genetics 129, 697–706.
Toda, T., Cameron, S., Sass, P., Zoller, M. and Wigler,
M. (1987). Three different genes in Saccharomyces
cerevisiae encode the catalytic subunits of the cyclic
AMP-dependent protein kinase. Cell 50, 277–288.
Tyers, M. and Futcher, B. (1993). Far1 and Fus3 link
the mating pheromone signal transduction pathway to
three G1-phase Cdc28 kinase complexes. Mol. Cell.
Biol. 13, 5659–5669.
Werner-Washburne, M., Braun, E., Johnston, G. C. and
Singer, R. A. (1993). Stationary phase in the yeast
Saccharomyces cerevisiae. Microbiol. Rev. 57, 383–
401.
Wrigglesworth, J. M. and Baum, H. (1980). The
biochemical functions of iron. In Jacobs, A. and
Worwood (Eds), Iron in Biochemistry and Medicine II.
Academic Press, New York.
Yamaguchi-Iwai, Y., Dancis, A. and Klausner, R. D.
(1995). AFT1: a mediator of iron regulated transcriptional control in Saccharomyces cerevisiae. EMBO J.
14, 1231–1239.
Yamaguchi-Iwai, Y., Stearman, R., Dancis, A. and
Klausner, R. D. (1996). Iron-regulated DNA binding
by the AFT1 protein controls the iron regulon in
yeast. EMBO J. 15, 3377–3384.
Zhang, Y. Z., Naleway, J. J., Larison, K. D., Huang, Z.
and Haugland, R. P. (1991). Detecting lacZ gene
expression in living cells with new lipophilic, fluorogenic â-galactosidase substrates. FASEB J. 5, 3108–
3113.
Zueco, J. and Boyd, A. (1992). A set of protein A fusion
vectors for use in combination with pEX vectors in
the production and affinity purification of specific
antibodies. Gene 121, 181–182.

. 13: 621–637 (1997)
Документ
Категория
Без категории
Просмотров
3
Размер файла
545 Кб
Теги
744
1/--страниц
Пожаловаться на содержимое документа