Downloaded from
`
`www.jbc.org
`
` by guest, on September 10, 2011
`
`THE JOURNAL OF BIOLOGICAL CHEMISTRY
`© 2003 by The American Society for Biochemistry and Molecular Biology, Inc.
`
`Vol. 278, No. 30, Issue of July 25, pp. 27636–27643, 2003
`Printed in U.S.A.
`
`Aft1p and Aft2p Mediate Iron-responsive Gene Expression in
`Yeast through Related Promoter Elements*□S
`
`Received for publication, January 5, 2003, and in revised form, May 2, 2003
`Published, JBC Papers in Press, May 19, 2003, DOI 10.1074/jbc.M300076200
`
`Julian C. Rutherford, Shulamit Jaron, and Dennis R. Winge‡
`From the Departments of Medicine and Biochemistry, University of Utah Health Sciences Center,
`Salt Lake City, Utah 84132
`
`The transcription factors Aft1p and Aft2p from Sac-
`charomyces cerevisiae regulate the expression of genes
`that are involved in iron homeostasis. In vitro studies
`have shown that both transcription factors bind to an
`iron-responsive element (FeRE) that is present in the
`upstream region of genes in the iron regulon. We have
`used DNA microarrays to distinguish the genes that are
`activated by Aft1p and Aft2p and to establish for the
`first time that each factor gives rise to a unique tran-
`scriptional profile due to the differential expression of
`individual iron-regulated genes. We also show that both
`Aft1p and Aft2p mediate the in vivo expression of FET3
`and FIT3 through a consensus FeRE. In addition, both
`proteins regulate MRS4 via a variant FeRE with Aft2p
`being the stronger activator from this particular ele-
`ment. Like other paralogous pairs of transcription fac-
`tors within S. cerevisiae, Aft1p and Aft2p are able to
`interact with the same promoter elements while main-
`taining specificity of gene activation.
`
`(5). Differential expression of Pdr1p/Pdr3p target genes may
`therefore involve different combinations of Pdr1p/Pdr3p at dif-
`ferent PDREs (5). Yap1p and Yap2p are 88% identical in their
`DNA binding regions and bind to the same consensus site (6).
`Analysis of global gene expression using microarrays has
`shown that although Yap1p and Yap2p are both involved in the
`response to cellular stresses, they regulate different regulons
`(7). The mechanism(s) of Yap1p/Yap2p gene selectivity are not
`understood but may include variations in the base pairs flank-
`ing the consensus Yap binding site (7).
`The iron regulon of S. cerevisiae is well characterized and
`includes genes that are involved in the uptake, compartmen-
`talization and use of iron. These include genes that encode
`siderophore transporters (ARN1–4), iron reductases (FRE1–6),
`iron permeases (FTR1, FET4), and multicopper oxidases that
`are involved in the coordinated oxidation and transport of iron
`across membranes (FET3, FET5) (8–12). In addition, three
`related genes (FIT1-FIT3) encode proteins that localize within
`the yeast cell wall and whose function is partially related to
`siderophore uptake (13).
`Genes within the iron regulon are induced under low iron
`conditions and are regulated by the transcription factors, Aft1p
`and Aft2p (14, 15, 18). Aft1p and Aft2p are 39% identical within
`their N-terminal regions, which contain the DNA binding do-
`main of each protein (15). The activation domain of Aft1p is
`within its C-terminal 413–572 amino acid residues which,
`when fused to the Gal4p DNA binding domain, activates tran-
`scription in an iron-independent manner (16). In addition, the
`N-terminal region of Aft1p contains a nuclear export signal
`that mediates differential cellular localization of Aft1p in re-
`sponse to iron (16). Aft1p is localized to the cytosol under
`iron-replete conditions, but is nuclear and active in iron-defi-
`cient conditions. Mutant alleles of AFT1 and AFT2 (AFT1-1up
`and AFT2-1up respectively) have been identified that result in
`the iron-independent activation of the iron regulon (14, 15).
`These gain of function mutations result from a single cysteine
`to phenylalanine substitution in identical regions of Aft1p and
`Aft2p.
`A strain that lacks a functional Aft1p (aft1⌬) grows poorly in
`iron-limiting conditions and exhibits reduced iron uptake and
`cell surface iron-reductase activity (14, 17). The vector-borne
`AFT2-1up allele is able to partially restore iron uptake in the
`aft1⌬ strain (15). The aft2⌬ strain has no growth phenotype on
`iron-limiting medium (15). However, the aft1⌬ aft2⌬ strain is
`more sensitive to iron-limiting conditions than the single aft1⌬
`strain (15, 18). The aft1⌬ strain also fails to grow with glycerol
`as a respiratory carbon source, and this phenotype is partially
`restored by complementation with low copy plasmids contain-
`ing the AFT2 or AFT2-1up alleles (15). Analysis of global gene
`expression has shown that the AFT2-1up allele activates the
`expression of a subset of the iron regulon (15). However, our
`previous microarray analysis did not compare transcript pro-
`27636
`
`Saccharomyces cerevisiae contains paralogous gene pairs
`that code for transcription factors that can have both overlap-
`ping and distinct functions. These transcription factors bind to
`the same promoter elements but generate distinct transcrip-
`tional profiles. Included in this group are Ace2p/Swi5p, Pdr1p/
`Pdr3p, and Yap1/Yap2p. Ace2p and Swi5p regulate the expres-
`sion of cell cycle-specific genes, are 83% identical in their zinc
`finger DNA-binding domains and recognize the same DNA
`binding site in vitro (1, 2). However, Ace2p and Swi5p can
`regulate separate genes, where discrimination between pro-
`moter elements is achieved through the interaction of these
`factors with other DNA-binding proteins (2). Alternatively, in
`cases where both transcription factors can induce the expres-
`sion of the same gene, one is often the more potent activator (3).
`There are genes that require both Ace2p and Swi5p for maxi-
`mal expression and others that are antagonistically regulated
`by these factors (4). Pdr1p and Pdr3p share 36% identity and
`regulate the expression of genes that are involved in pleiotropic
`drug resistance. Both these factors are able to bind to the same
`DNA element (PDRE) in vivo as either hetero- or homodimers
`
`* This research was supported by Grant CA 61286 from NCI, Na-
`tional Institutes of Health (to D. R. W.) and Microarray Supplemental
`Award ES03817 from the National Institutes of Environmental Health
`Sciences (to D. R. W.). The costs of publication of this article were
`defrayed in part by the payment of page charges. This article must
`therefore be hereby marked “advertisement” in accordance with 18
`U.S.C. Section 1734 solely to indicate this fact.
`□S The on-line version of this article (available at http://www.jbc.org)
`contains EXCEL spreadsheets containing the normalized, filtered, and
`averaged data.
`‡ To whom correspondence should be addressed. Tel.: 801-585-5103;
`Fax: 801-585-5469; E-mail: dennis.winge@hsc.utah.edu.
`
`This paper is available on line at http://www.jbc.org
`
`

`
`Downloaded from
`
`www.jbc.org
`
` by guest, on September 10, 2011
`
`Regulation of Expression by Aft1p and Aft2p
`
`27637
`
`files for both AFT2-1up and AFT1-1up cells. Additionally, the
`AFT1-1up and the AFT2-1up alleles differentially regulate the
`expression of two iron-regulated genes (MRS4 and FIT2) (15).
`Aft1p mediates transcriptional regulation through an iron-
`responsive element (FeRE)1 that has the consensus sequence
`PyPuCACCCPu (14). The FeRE was identified by the in vivo
`analysis of the FET3 promoter region using lacZ fusion con-
`structs and DNA footprinting (14). Aft1p binds to the FET3
`FeRE in vivo in an iron-dependent manner. Sequences similar
`to the FET3 FeRE were identified in the promoter regions of
`five other known iron-regulated genes. In vitro assays con-
`firmed the binding of Aft1p to these particular sequences, from
`which the consensus FeRE sequence was derived (14). SMF3,
`that encodes a metal transporter, is also iron-regulated in an
`Aft1p/Aft2p-dependent manner through a consensus FeRE
`(19). N-terminal truncates of Aft1p and Aft2p bind to the FET3
`FeRE in vitro, consistent with both factors regulating the iron
`regulon through the consensus FeRE (15). Expression of AFT2
`from a high copy plasmid, but not the chromosomal copy of
`AFT2, activates the expression of a lacZ reporter under the
`control of the FET3 FeRE in an iron-dependent manner (18).
`However, the high expression of one of a pair of transcription
`factors can result in the aberrant activation of genes that are
`specifically regulated by the other factor (2). It is therefore not
`clear to what extent Aft1p and Aft2p interact with the same
`responsive element in vivo.
`The evidence is therefore consistent with Aft1p and Aft2p
`having overlapping but non-redundant roles in the transcrip-
`tional regulation of the iron regulon in S. cerevisiae . We are
`interested in understanding the selective advantage of this
`organism having two iron-responsive transcription factors. In
`this study, microarray experiments have been used to further
`define the activities of the gain of function mutants and dem-
`onstrate that they give rise to different global transcriptional
`profiles. We have also analyzed the ability of Aft1p and Aft2p to
`activate transcription through the same consensus FeRE. In
`vivo lacZ reporter constructs show clearly that Aft1p and Aft2p
`activate gene expression through the consensus FeREs in the
`FET3 and FIT3 promoter regions. Both Aft1p and Aft2p also
`induce the activation of MRS4 through a variant FeRE. An
`extended FeRE has been identified that is overrepresented in
`the genes that are most highly induced by the AFT1-1up and
`AFT2-1up alleles.
`
`MATERIALS AND METHODS
`Yeast Strains and Culture Conditions—Haploid aft1⌬aft2⌬ strains
`were isolated following the mating of the strains, BY4741aft2⌬ (MATa
`his3⌬1 leu2⌬0 met15⌬0 ura3⌬0 aft2::kanMX4) and BY4742aft1⌬
`(MAT␣ his3⌬1 leu2⌬0 lys2⌬0 ura3⌬0 aft1::kanMX4), which were pur-
`chased from Research Genetics. A strain containing a copy of the
`AFT2-1up allele integrated at the HIS3 locus was generated by trans-
`forming a haploid aft1⌬aft2⌬ strain (MAT␣ his3⌬1 leu2⌬0 ura3⌬0
`aft1::kanMX4 aft2::kanMX4) with plasmid pAFT2-1upINT that had
`been linearized using MscI and selecting for growth on agar plates
`lacking histidine. The allele status of the aft1⌬aft2⌬ strains and the
`integration of the AFT2-1up allele at the HIS3 locus were verified using
`PCR and DNA sequencing. Cells were grown with 2% glucose as a
`carbon source, in 1% yeast extract, 2% peptone medium (YPD), com-
`plete-synthetic medium (CMD) or, when appropriate, complete-syn-
`thetic medium lacking uracil and/or histidine (CMD-Ura, CMD-His,
`CMD-Ura/His).
`Vectors and Fusion Genes—The YCp plasmids, pAFT1-1up and
`pAFT2-1up are HIS3 derivatives of pAFT1-1up and pAFT2-1up (15) from
`which the AFT1 (XhoI/SacI) and AFT2 (BamHI/XbaI) sequences have
`been subcloned into the YCp plasmid, pRS413. Consequently, the gain
`of function mutant alleles of AFT1 and AFT2 within pAFT1-1up and
`
`1 The abbreviations used are: FeRE, iron-responsive element; ORF,
`open reading frame; CMD, complete-synthetic medium; UAS, upstream
`activation sequence.
`
`pAFT2-1up are under the control of their own promoters (15). The
`integrative vector pAFT2-1upINT was generated by subcloning the
`AFT2 sequences (XhoI/XbaI) from pAFT2-1up into the YIp plasmid,
`pRS403. The lacZ reporter constructs, pFC-W and pFC-LM2 were a gift
`from Andrew Dancis (14). These contain, respectively, a functional and
`non-functional copy of the FET3 FeRE in a minimal promoter. The
`plasmid pFET3-lacZ (0.5 kb of the 5⬘-region of FET3, inclusive of the
`start codon, fused to the lacZ gene in YEp354, Ref. 20) was a gift from
`Jerry Kaplan (21). To generate a version of pFET3-lacZ in which the
`same FeRE that is present in pFC-W was mutated to a non-functional
`FeRE (CACCC to CAGGG), pFET3-lacZ was used as template for
`QuikChange mutagenesis (Stratagene) using the primer 5⬘-GGCAAG-
`GCCCATCTTCAAAAGTGCAGGGATTTGCAGGTGCTCTTATTCTCG-
`CC-3⬘ and its complement to generate pFET3-lacZ (mut). The 0.5-kb
`fragment containing the FET3 sequences from pFET3-lacZ and pFET3-
`lacZ (mut) were excised and ligated into the PstI/SmaI sites of Yep354
`to generate pFET3-lacZ and pFET3-lacZ (mut) respectively. To gener-
`ate pFIT3-lacZ (1 kb of 5⬘-region of FIT3, inclusive of the start codon,
`fused to the lacZ gene in YEp354) yeast-genomic DNA from strain
`BY4741 was isolated and used as template for PCR using the primers
`5⬘-CGGGATCCTCCATAAACATTTCCTTTGTC-3⬘ and 5⬘-CCAAGCTT-
`TCATTTTAGGGATTATTGTTATTAG-3⬘. The resulting 1-kb fragment
`was ligated into the BamHI/HindIII sites of Yep354 to generate
`pSJDW36. Plasmid pSJDW36 was used as template for QuikChange
`mutagenesis using the primer 5⬘-CATCAAAAAATATGGGATAGCG-
`CCCTGCGCAACAAACACCCTGCAAAAAAAAATCTAGGACATAGG-3⬘
`and its complement to mutate two potential FeREs to non-functional
`FeREs (CACCC to CAGGG) and thereby generate pSJDW36 (mut). The
`FIT3 sequences from both pSJDW36 and pSJDW36 (mut) were excised
`and ligated to the BamHI/HindIII sites of Yep354 to generate pFIT3-
`lacZ and pFIT3-lacZ (mut), respectively. To generate plasmid pMRS4a-
`lacZ (1 kb of 5⬘-region of MRS4, inclusive of the start codon, fused to the
`lacZ gene in YEp354) yeast-genomic DNA from strain BY4741 was used
`as template for PCR using the primers 5⬘- CGCACAGGATCCTCGAA-
`GATAGCGTAGCGTTC-3⬘ and 5⬘-GGGAGCAAGCTTCATAATATTAA-
`CTGATATTTCGGTTG-3⬘ (primer Mrs4a). The resulting 1-kb fragment
`was ligated into the BamHI/HindIII sites of pHOLLY to generate
`pJRDW25. Plasmid pJRDW25 was used as template for PCR using
`primer Mrs4a with, individually, primers 5⬘-CGCACAGGATCCCATA-
`TTTGGAATTCAGC-3⬘, 5⬘-CGCACAGGATCCCCCACAGGAATCGCT-
`AC-3⬘ , 5⬘-CGCACAGGATCCCGTGTCTCTTTTCGGTA-3⬘ , 5⬘-CGCAC-
`AGGATCCGAATGAGAGCATGGCGA-3⬘, and 5⬘-CGCACAGGATCCC-
`TTTTGCCTACCATTGG-3⬘ to generate sequential truncations of the
`MRS4 promoter region. The resulting fragments, and the 1-kb BamHI/
`HindIII fragment from pJRDW25, were ligated into the BamHI/Hind-
`III sites of Yep354 to generate plasmids pMRS4lacZ(a-f). Plasmid
`pMRS4-lacZd was used as template for QuikChange mutagenesis using
`the primer 5⬘-CGAAGACTGAAAGGCAAGAACAGGGTGCTATCTTTT-
`GCCTACCATTG-3⬘ and its complement and primer 5⬘-GTCTCTTTT-
`CGGTATTTTGGCAGGGTTTCTTGAATGAGAGCATGGC -3⬘ and its
`complement to generate, respectively, plasmids, pMRS4-lacZd (mut1)
`and pMRS4-lacZd (mut2) that contain CACCC to CAGGG mutations in
`individual FeRE-like sequences. To generate a reporter construct that
`contains the FeRE of MRS4 within the ⌬UAS CYC1 promoter fused to
`the lacZ gene, 200 pmol of each of the partial overlapping oligos 5⬘-tc-
`gaAAGAAAGGGTGCCAAAA-3⬘ and 5⬘-ctagTTTTGGCACCCTTTC-
`TT-3⬘ were mixed in SSC (final 0.32⫻), boiled for 15 min, and then
`incubated overnight at 55 °C. The annealed products were then ligated
`into the XhoI/XbaI sites of pNB404 (22) to generate pMRS4-FeRE. All
`PCR-generated sequences were confirmed by DNA sequencing. All
`yeast transformations were performed using the lithium acetate proce-
`dure. In the case of transformations using aft1⌬aft2⌬ strains, cells were
`pregrown in YPD under nitrogen and agar plates supplemented with
`FeCl2 (100 ␮M).
`mRNA Quantification by S1 Nuclease Analysis—Strain aft1⌬aft2⌬
`(MAT ␣ his3⌬1 leu2⌬0 ura3⌬0 aft1::kanMX4 aft2::kanMX4) was used
`for transformation of all lacZ reporter constructs with separately
`pAFT1-1up, pAFT2-1up or pRS413. Cells were harvested at mid-log
`phase, and total RNA was extracted using the hot acidic phenol method
`(31). DNA oligonucleotides, with sequences complementary to lacZ,
`FIT3 and CMD1 mRNA (calmodulin as an internal loading control),
`were end-labeled with 32P using T4 polynucleotide kinase (New Eng-
`land Biolabs). S1 analysis was carried out as previously described (1).
`Briefly, the 32P-labeled oligonucleotides were hybridized with 12 ␮g of
`total RNA in HEPES buffer (38 mM HEPES, pH 7.0, 0.3 M NaCl, 1 mM
`EDTA, 0.1% Triton X-100) at 55 °C overnight. The reactions were then
`treated with 35 units of S1 nuclease (Promega), and the resulting
`DNA:RNA double-stranded duplexes were ethanol-precipitated and
`
`

`
`Downloaded from
`
`www.jbc.org
`
` by guest, on September 10, 2011
`
`27638
`
`Regulation of Expression by Aft1p and Aft2p
`
`TABLE I
`Genes induced by AFT1–1up
`Microarray analysis was carried out comparing transcript levels of control aft1⌬aft2⌬ cells and cells containing pAFT1–1up. The 25 most highly
`induced genes are shown. The mean and S.D. of the Cy5/Cy3 ratios of three independent experiments are listed. Also shown are the equivalent data
`(mean, S.D.) from the experiments involving pAFT2–1up. Listed is the number of consensus FeRE sites (PyPuCACCC) within 1 kb of the start codon
`of each gene. Notes: (1) AFT1 is not present in the control strain. (2) YDR271C overlaps with CCC2 - a known iron regulated gene. (3) A transcript
`is not detected (ND) when the signal is not greater than the background signal.
`
`Gene name
`
`Mean
`
`FIT3
`FIT2
`FET3
`FIT1
`YOR387C
`ENB1 (ARN4)
`SIT1 (ARN3)
`TIS11
`FTR1
`AFT1 (1)
`FRE2
`FRE1
`TAF1 (ARN2)
`FRE3
`YHL035C
`FRE5
`BIO5
`YDR271C (2)
`ARN1
`FRE6
`ATX1
`OYE3
`COT1
`YMR034C
`AKR1
`a ND, not detected.
`
`43.4
`30.9
`25.2
`8.7
`6.7
`6.7
`6.3
`5.1
`5.0
`4.8
`4.1
`4.0
`3.4
`3.0
`3.0
`2.9
`2.8
`2.6
`2.5
`2.4
`2.4
`2.3
`2.3
`2.2
`2.2
`
`S.D.
`
`8.6
`11.0
`12.3
`2.0
`0.6
`1.0
`2.2
`2.0
`0.5
`2.0
`0.6
`0.7
`0.6
`0.5
`0.5
`0.4
`0.3
`0.2
`0.8
`0.2
`0.2
`0.3
`0.2
`0.4
`0.1
`
`AFT2–1up
`
`S.D.
`
`# FeRE
`
`14.6
`2.4
`3.7
`14.3
`2.0
`2.0
`1.9
`3.6
`2.6
`NDa
`ND
`3.5
`1.1
`2.0
`2.2
`1.9
`1.7
`1.9
`1.6
`2.2
`2.0
`2.7
`2.3
`2.2
`1.5
`
`2.1
`0.2
`0.2
`1.8
`0.7
`0.7
`0.3
`1.0
`0.2
`
`0.6
`0.1
`0.4
`0.2
`0.3
`0.2
`0.5
`0.2
`0.2
`0.6
`0.4
`0.2
`0.1
`0.1
`
`3
`5
`2
`3
`2
`3
`1
`2
`2
`1
`2
`1
`1
`3
`1
`3
`0
`0
`1
`2
`1
`0
`1
`1
`1
`
`heat-denatured in formamide buffer. The samples were then separated
`using an 8% polyacrylamide/8.3 M urea gel.
`Microarray Analysis—Strain aft1⌬aft2⌬ (MATa his3⌬1 leu2⌬0
`lys2⌬0 met15⌬0 ura3⌬0 aft1::kanMX4 aft2::kanMX4) was transformed
`with separately, pAFT1-1up, pAFT2-1up and their corresponding paren-
`tal vectors as controls (pRS316, pRS416). Cells were grown in 300 ml of
`CMD-Ura medium and harvested at an OD600 nm of 0.4. Total RNA was
`isolated by the hot acidic phenol method and mRNA was isolated from
`total RNA using the PolyATtract mRNA Isolation System IV kit (Pro-
`mega) following the manufacturer’s instructions. mRNA (1.2 ␮g) was
`used to generate cDNA probes by reverse transcription (Superscript II,
`GIBCO) with incorporation of Cy3-dCTP or Cy5-dCTP (Amersham Bio-
`sciences). The arrayed slides were produced as follows. The PCR am-
`plification products of all the S. cerevisiae open reading frames (ORFs)
`were purchased from Research Genetics. Each PCR product was ream-
`plified and purified on glass fiber filter plates (Millipore). The reampli-
`fied products representing ⬃6000 ORFs (4% not represented) were
`diluted to 50% Me2SO and spotted in duplicate on 3-aminopropyl-
`methyldiethoxy silane-coated slides using a Generation III Microarray
`Spotter (Amersham Biosciences). The cDNA-labeled probes were then
`hybridized onto an arrayed slide and fluorescence was captured with a
`GEN III dual-laser confocal scanner (Molecular Dynamics). Fluorescent
`intensities were quantified using Arrayvision 6.0 (Imaging Research).
`The ratios of signal intensities (Cy5/Cy3) were normalized by the me-
`dian output signal intensity for all genes, and the normalized data for
`three independent experiments were averaged. The data were filtered
`to remove those measurements that were not higher than the slides’
`background fluorescence and assembled into EXCEL spreadsheets.
`Analysis of the potential regulatory sequences within the most highly
`induced genes was carried out using the program GeneSpring 3.2 (Sil-
`icon Genetics). The EXCEL spreadsheets containing the normalized,
`filtered, and averaged data are available as Supplemental Material.
`
`RESULTS
`Microarray Analysis of Aft1p and Aft2p Regulons—Previous
`studies have shown that Aft1p and Aft2p differentially activate
`the expression of a subset of the iron-regulated genes in S.
`cerevisiae (15). We were interested to learn the effect of Aft1p
`and Aft2p on global gene expression in cells of the same genetic
`background. DNA microarray analysis was initiated to com-
`pare the expression profile resulting from the AFT1-1up allele
`
`and, separately, the AFT2-1up allele. The aft1⌬aft2⌬ strain was
`used into which pAFT1-1up, pAFT2-1up and control vectors had
`been separately transformed. The AFT1-1up/AFT2-1up alleles
`were expressed from their own promoters in low copy plasmids,
`so as to minimize the overexpression of the gain of function
`alleles. The transcriptional profiles of the pAFT1-1up- and the
`pAFT2-1up-containing strains were separately compared with
`that of control strains. Each experiment was carried out inde-
`pendently on three occasions. The average induction ratio for
`each gene was calculated, and the genes that were most highly
`induced in cells containing the AFT1-1up and AFT2-1up alleles
`were identified (Tables I and II, Fig. 1). These data will include
`genes that are the direct targets of Aft1p and/or Aft2p or genes
`that are activated as the result of the indirect effects of Aft1p/
`Aft2p. A complete data set can be found in the Supplemental
`Material.
`Many of the genes identified as being induced as the result of
`the AFT1-1up/AFT2-1up alleles are known to code for proteins
`that are involved in iron metabolism. Some of these genes are
`induced to a similar extent as the result of either Aft1p or Aft2p
`(e.g. FRE6, ATX1, COT1). However, the more highly induced
`genes tend to be activated to a greater extent when Aft1p is
`present (e.g. FIT3, FIT2, FTR1). There are examples of genes
`that contain a consensus FeRE but which are only induced as
`the result of one of the Aft1p/Aft2p factors. These genes include
`the Aft1p-specific FRE2 and ARN2 and the Aft2p-specific
`BNA2 and ECM4. BNA2 codes for a protein with similarity to
`indoleamine 2,3-dioxygenases and Ecm4p is involved in cell
`wall biosynthesis, although its precise function is not known.
`Included in the group of genes that are preferentially induced
`as the result of Aft1p is YOR387C. Interestingly, YOR387C is
`located within a region of chromosome XV that contains a
`group of genes (FIT2, FIT3, FRE3) that are highly induced in
`cells containing either AFT1-1up or AFT2-1up. YOR387C codes
`for a 206-amino acid protein that shares 93% identity with the
`mitochondrial Vel1p from S. cerevisiae (23). Common to both
`
`

`
`Regulation of Expression by Aft1p and Aft2p
`
`27639
`
`TABLE II
`Genes induced by AFT2–1up
`Microarray analysis was carried out comparing transcript levels of control aft1⌬aft2⌬ cells and cells containing pAFT2–1up. The 25 most highly
`induced genes are shown. The mean and the S.D. of the Cy5/Cy3 ratios of three independent experiments are listed. Also shown are the equivalent
`data (mean, S.D.) from the experiments involving pAFT1–1up. Listed is the number of consensus FeRE sites (PyPuCACCC) within 1 kb of the start
`codon of each gene. Notes: (1) AFT2 is not present in the control strain (2) YOR225w overlaps with ISU2- a known iron regulated gene (3). A
`transcript is not detected (ND) when the signal is not greater than the background signal.
`AFT1–1up
`
`S.D.
`
`# FeRE
`
`Gene name
`
`Mean
`
`S.D.
`
`Downloaded from
`
`www.jbc.org
`
` by guest, on September 10, 2011
`
`FIT3
`FIT1
`AFT2 (1)
`FET3
`TIS11
`FRE1
`BNA2
`ECM4
`YAP2
`YKR104W
`OYE3
`SMF3
`FTR1
`YOL083W
`MRS4
`FIT2
`YMR041C
`YBR012C
`LAP4
`YOR225W (2)
`HSP26
`COT1
`YGR146C
`ISU1
`YHL035C
`a ND, not detected.
`
`14.6
`14.3
`4.4
`3.7
`3.6
`3.5
`3.0
`2.8
`2.8
`2.7
`2.7
`2.6
`2.6
`2.5
`2.5
`2.4
`2.4
`2.4
`2.3
`2.3
`2.3
`2.3
`2.3
`2.2
`2.2
`
`2.1
`1.8
`1.2
`0.2
`1.0
`0.6
`0.6
`0.5
`0.2
`0.2
`0.4
`0.2
`0.2
`0.3
`0.3
`0.2
`0.2
`0.2
`0.4
`0.5
`0.6
`0.2
`0.2
`0.4
`0.2
`
`43.4
`8.7
`NDa
`25.2
`5.1
`4.0
`ND
`ND
`1.7
`ND
`2.3
`1.9
`5.0
`1.8
`1.7
`30.9
`1.5
`ND
`1.6
`1.7
`1.5
`2.3
`1.5
`1.5
`3.0
`
`8.6
`2.0
`
`12.3
`2.0
`0.7
`
`0.2
`
`0.3
`0.1
`0.5
`0.2
`0.4
`11.0
`0.4
`
`0.1
`0.2
`0.4
`0.2
`0.1
`0.3
`0.5
`
`3
`3
`0
`2
`2
`1
`1
`1
`0
`0
`0
`2
`2
`1
`0
`5
`0
`0
`1
`0
`0
`1
`1
`1
`1
`
`FIG. 1. Summary of the genes induced by Aft1p and Aft2p. The
`overlap represents those genes that are induced by both Aft1p and
`Aft2p. Those genes that are known to encode proteins that are involved
`in iron homeostasis are shaded. AFT1 and AFT2 are excluded.
`
`experiments are OYE3 and YHL035C, which have not previ-
`ously been identified as targets of the Aft1p/Aft2p factors.
`OYE3 codes for a NADPH dehydrogenase and does not contain
`a consensus FeRE within its 5⬘-upstream region and conse-
`quently may be induced as an indirect result of the Aft1p/Aft2p
`factors. YHL035C has one consensus FeRE within its 5⬘-up-
`stream region and codes for a member of the ATP-binding
`cassette (ABC) family of transporters, however its localization
`and function are unknown.
`Integration of AFT2-1up Allele—To confirm that the use of
`plasmid-borne constitutive alleles for the microarray experi-
`ments was an appropriate strategy, the AFT2-1up locus, under
`the control of its own promoter and terminator, was integrated
`into the HIS3 locus of the aft1⌬aft2⌬ strain. S1 nuclease pro-
`tection assays were used to analyze the expression of FIT3 in
`the aft1⌬aft2⌬ strain containing a control plasmid or pAFT2-
`1up, the aft1⌬aft2⌬ strain with the integrated AFT2-1up allele
`and the aft1⌬ strain. The expression of FIT3 that resulted from
`
`FIG. 2. Comparison of activation of the expression of FIT3 by
`the chromosomal and plasmid-borne AFT2-1up allele. S1 nuclease
`protection assays were used to quantify mRNA level of FIT3. RNA was
`isolated from the aft1⌬aft2⌬ strain transformed with a control plasmid
`or pAFT2-1up and the aft1⌬aft2⌬ strain into which the AFT2-1up allele
`had been integrated into the HIS3 locus (aft1⌬AFT2-1up). Each strain
`was grown in CMD (-His) medium with or without added iron (100 ␮M)
`to mid-log phase. RNA was also isolated from the aft1⌬ strain that had
`been grown in CMD with and without the iron chelator BPS (100 ␮M).
`The upper band for each sample is the FIT3 gene, and the lower band
`is the calmodulin-loading control (CMD1).
`
`the plasmid-borne AFT2-1up allele was only slightly greater
`than that of the integrated AFT2-1up allele (Fig. 2). Conse-
`quently, it is unlikely that the microarray data include genes
`that have been artificially induced through overexpression of
`the AFT1-1up and AFT2-1up alleles. Interestingly, the AFT2-
`1up allele still responds to iron to some extent when it is present
`in the chromosome, but that response is lost in the strain
`containing the plasmid-borne AFT2-1up allele.
`In Vivo Analysis of FET3 FeRE—Truncated Aft1p and Aft2p
`that contain the N-terminal basic regions of each protein, are
`able to bind to the FET3 FeRE in vitro (15). As both the
`AFT1-1up and AFT2-1up alleles activate the expression of many
`of the same genes, we initially used two reporter constructs to
`learn if both factors act through the same FeRE in vivo. The
`first reporter construct contains 500 bp of the FET3 5⬘-up-
`stream region, which has 3 FeRE-like sequences, fused to the
`
`

`
`Downloaded from
`
`www.jbc.org
`
` by guest, on September 10, 2011
`
`27640
`
`Regulation of Expression by Aft1p and Aft2p
`
`FIG. 3. Aft1p and Aft2p regulate the expression of FET3
`through the same FeRE. A, the nucleotide sequences of 3 FeRE-like
`elements within 500 bp of the start codon of FET3. Nucleotides that
`deviate from the consensus FeRE sequence identified in Ref. 14 are
`underlined. B, plasmid pFC-W contains the ⫺251 FeRE that has been
`inserted into the CYC1 promoter and fused to the lacZ gene. Plasmid
`pFET3-lacZ contains 500 bp of the upstream region of FET3 fused to
`the lacZ gene. Ovals represent FeRE-like sequences (filled in) or mu-
`tated FeREs (hatched). C, S1 nuclease protection assays were used to
`quantify mRNA levels of lacZ. RNA was isolated from the aft1⌬aft2⌬
`strain that had been separately transformed with pRS413 (control),
`pAFT1-1up and pAFT2-1up together with the lacZ reporter constructs
`pFC or pFET3-lacZ and their respective control plasmids [pFC-W and
`pFET3-lacZ (mut)]. Each strain was grown in CMD (-His/Ura) to mid-
`log phase. The upper band for each sample is the lacZ gene. and the
`lower band is the calmodulin-loading control (CMD1).
`
`lacZ reporter gene (pFET3-lacZ)(Fig. 3B). A mutated fusion
`gene was generated in which the core CACCC sequence of the
`FeRE closest to the lacZ start codon was changed to the se-
`quence, CAGGG. This mutation is known to create a non-
`functional FeRE (14). The second reporter construct consists of
`the same FET3 single FeRE within a minimal CYC1 promoter/
`lacZ reporter construct that lacks an upstream activating se-
`quence (pFC-W). The corresponding control vector contains a
`non-functional CAGGG sequence within the FeRE (pFC-LM2)
`(14). Plasmids pFET-lacZ and pFC-W, and their corresponding
`control vectors, were
`transformed separately into
`the
`aft1⌬aft2⌬ strain containing either pAFT1-1up or pAFT2-1up.
`S1 nuclease protection assays were used to analyze the expres-
`sion of the lacZ reporter genes. Both the AFT1-1up and AFT2-
`1up alleles induced the expression of lacZ in pFET-lacZ and
`pFC-W, but not the control reporter genes containing a non-
`functional FeRE (Fig. 3C). In both cases, the AFT2-1up induced
`expression of the reporter genes was weaker than the activa-
`tion by the AFT1-1up allele. The AFT2-1up allele is a strong
`activator of the chromosomal FET3 (15), so the minimal effects
`of this allele with these particular reporter constructs may be
`due to the lack of upstream sequences of the FET3 promoter
`region that are absent in pFET3-lacZ and pFC-W.
`In Vivo Analysis of FIT3 FeRE—The most highly induced
`gene by the AFT1-1up and AFT2-1up alleles is FIT3 (Tables I
`and II). To analyze the extent that both Aft1p and Aft2p acti-
`vate the expression of FIT3 through the same FeRE, we gen-
`erated a reporter construct, containing 1 kb of the FIT3 5⬘-
`region fused to lacZ (pFIT3-lacZ). It was anticipated that the
`larger 5⬘-upstream region would be representative of the AFT2-
`1up induced expression of the chromosomal copy of FIT3. There
`are 4 potential FeRE sites within the 1-kb 5⬘-region of FIT3
`
`FIG. 4. Aft1p and Aft2p regulate the expression of FIT3
`through the same FeRE. A, the nucleotide sequences of 4 FeRE-like
`elements within 1 kb of the start codon of FIT3. Nucleotides that
`deviate from the consensus FeRE sequence identified in Ref. 14 are
`underlined. B, plasmid pFET3-lacZ contains 1 kb of the upstream
`region of FIT3 fused to the lacZ gene. Ovals represent FeRE-like se-
`quences (filled in) or mutated FeREs (hatched). C, S1 nuclease protec-
`tion assays were used to quantify mRNA levels of lacZ. RNA was
`isolated from the aft1⌬aft2⌬ strain that had been separately trans-
`formed with pRS413 (control), pAFT1-1up and pAFT2-1up together with
`the lacZ reporter constructs pFIT3-lacZ and its control pFIT3-lacZ
`(mut). Each strain was grown in CMD (-His/Ura) to mid-log phase. The
`upper band for each sample is the lacZ gene, and the lower band is the
`calmodulin-loading control (CMD1).
`
`that each deviate from the consensus FeRE by one nucleotide
`(Fig. 4A). To see if loss of an FeRE would have the same effect
`for both Aft1p- and Aft2p-dependent activation, we generated a
`FIT3-lacZ fusion gene that contained mutations in two of these
`potential FeRE sites. The core CACCC FeRE sequence was
`changed to the non-functional CAGGG (Fig. 4B). Plasmid
`pFIT3-lacZ and its control were transformed separately into
`the aft1⌬aft2⌬ strain that contained either pAFT1-1up or
`pAFT2-1up. Disruption of the two FeRE sites attenuates both
`the AFT1-1up and AFT2-1up induced expression of the FIT3-
`lacZ fusion (Fig. 4C). Both of these FeRE sites deviate from the
`consensus FeRE in their most 3⬘-nucleotide, consistent with
`this nucleotide being unimportant for recognition by Aft1p and
`Aft2p. In agreement with the microarray data, the AFT1-1up
`allele is a stronger activator than the AFT2-1up allele of the
`expression of FIT3. Together, the analysis of the expression of
`the FET3-lacZ and FIT3-lacZ fusion genes is consistent with
`the AFT1-1up and AFT2-1up alleles inducing expression
`through the same core FeRE.
`In Vivo Analysis of MRS4 FeRE—We have previously shown
`that MRS4 is preferentially induced by the AFT2-1up allele
`(15). Interestingly, the 1-kb 5⬘-upstream region of MRS4 con-
`tains 5 FeRE-like sequences that each deviate from the con-
`sensus FeRE by 2 nucleotides (Fig. 5A). To determine whether
`the AFT1-1up and AFT2-1up alleles activate the expression of
`MRS4 through these non-consensus FeRE sites, a series of
`MRS4-lacZ fusions were generated which contain truncations
`of the MRS4 upstream region (Fig. 5B). These reporter con-
`structs were transformed separately into the aft1⌬aft2⌬ strain
`containing either pAFT1-1up or pAFT2-1up. The presence of at
`least two putative FeRE-like sites resulted in AFT2-1up-de-
`pendent lacZ expression, whereas the fusion with only one
`putative FeRE-like site did not (Fig. 5C). Similar results were
`observed in cells that contained pAFT1-1up (Fig. 5C). Consist-
`
`

`
`Downloaded from
`
`www.jbc.org
`
` by guest, on September 10, 20