(Received for publication, July 30, 1996, and in revised form, October 24, 1996)
From the Division of Immunology and Cell Biology, Department of Pathology, University of Utah School of Medicine, Salt Lake City, Utah 84132
Genetic studies have demonstrated that high affinity ferrous transport in Saccharomyces cerevisiae requires an oxidase (Fet3p) and a permease (Ftr1p). Using an iron-independent galactose-based expression system, we show that expression of these two genes can mediate high affinity ferrous iron transport, indicating that these two genes are not only necessary, but sufficient for high affinity iron transport. Schizosaccharomyces pombe also employ an oxidase-permease system for high affinity iron transport. The S. pombe genes, fio1+ (ferrous iron oxidase) and fip1+ (ferriferous permease), share significant similarity to FET3 and FTR1 from S. cerevisiae. Both fio1+ and fip1+ are transcriptionally regulated by iron need, and disruption of fio1+ results in a loss of high affinity iron transport. Expression of fio1+ alone in an S. cerevisiae fet3 disruption strain does not result in high affinity iron transport. This result indicates that the S. pombe ferroxidase, while functionally homologous to the S. cerevisiae ferroxidase, does not have enough similarity to interact with the S. cerevisiae permease. Simultaneous expression of both S. pombe genes, fio1+ and fip1+, in S. cerevisiae can reconstitute high affinity iron transport. These results demonstrate that the oxidase and permease are all that is required to reconstitute high affinity iron transport and suggest that such transport systems are found in other eukaryotes.
Saccharomyces cerevisiae has two plasma membrane transporters for elemental iron, a high affinity system (Km = 0.15 µM) that transports only iron and a low affinity system (Km = 30 µM) that is also capable of transporting other metals such as cobalt and cadmium (1, 2). Both systems rely on cell surface ferrireductases to convert extracellular ferric to ferrous iron. The low affinity transport system is encoded by the FET41 gene (2). Two genes, FET3 and FTR1, are directly required for high affinity iron transport (3, 4). The FET3-FTR1 based transport system consists of an oxidase (Fet3p) and a permease (Ftr1p) that work together to facilitate transmembrane iron transport. The Fet3p is a type 1 plasma membrane protein in which the multicopper oxidase domain responsible for ferroxidase activity is present on the extracellular surface (5). Defects in copper transport result in reduced iron transport by affecting the production of active Fet3p (6, 7). The permease, Ftr1p, has multiple membrane spanning domains and contains a motif, REGLE, that may be involved in iron binding (4). Simultaneous expression of Fet3p and Ftr1p is required for proper localization of both proteins at the cell surface, suggesting that a complex is formed during movement of the proteins through the secretory pathway.
Genetic studies have demonstrated that FET3 and FTR1 are necessary for high affinity transport, but it has not been established that the combination of Fet3p and Ftr1p alone is sufficient to reconstitute transport activity. Normally FET3 and FTR1 are regulated transcriptionally by iron need and iron transport is not induced in high iron medium (1, 3, 4, 8). When incubated in low iron medium, cells with high-copy plasmids containing the FET3 and FTR1 genes show supernormal levels of iron transport activity (4). In this study, we demonstrate that expression of both genes from an iron-independent promoter results in high affinity iron transport even when the cells are grown in iron-replete conditions.
A transport system that relies on the permease-oxidase complex has been characterized only in the budding yeast S. cerevisiae. Schizosaccharomyces pombe also requires a cell surface ferrireductase to utilize ferric iron (9). This reductase has limited amino acid similarity to the S. cerevisiae ferrireductase FRE1 and displays more similarity to the mammalian gp91-phox subunit of the human NADPH phagocyte oxidoreductase. Examination of the DNA data bases revealed S. pombe genes very similar to FET3 and FTR1. In this report, we describe studies that identify these genes as the molecular components responsible for the S. pombe high affinity inducible iron transport system. Expression of these two S. pombe genes in S. cerevisiae reconstitutes high affinity iron transport, indicating that the S. pombe genes, like the S. cerevisiae genes, are necessary and sufficient for iron transport.
The S. cerevisiae strains used
in this study were the wild-type F113 (10) and the fet3
disruption strain 1397-6A (3). The S. pombe strains used
were the wild-type FY254 (h, can1-1, leu1-32,
ade6-M210, ura4-D18) and the fio1 disruption strain 4051. This strain was generated by creating a fio1 disruption construct (fio1::URA4) by digesting the plasmid GALFIO1 (see
below) with EcoRI, treating the vector with Klenow, and
ligating the blunted HindIII fragment of pTZURA4 containing
the ura4+ gene (obtained from Dr. Susan Forsburg
at the Salk Institute, San Diego, CA). The fio1 disruption
construct was cut with PvuII and HindIII, and the
fragment containing the disrupted fio1 gene was purified and
transformed into FY254 using the standard lithium acetate
transformation protocol for S. pombe (11). Colonies able to
grow on medium lacking uracil were examined by Southern analysis for
the presence of the disrupted allele (data not shown). Strain 4051 was
identified as containing the URA4 gene within fio1+.
The media used in this study included YPD (1.0% yeast extract, 0.2% peptone, 2.0% glucose) and CM-URA-LEU (0.69% yeast nitrogen base, 0.13% amino acid supplementation excluding uracil and leucine) with either 2.0% glucose or 2.0% galactose as the carbon source. Iron-limited medium was generated by adding bathophenanthrolinedisulfonic acid (BPS),2 disodium salt hydrate from Aldrich to a final concentration of 80 µM in YPD (BPS(0)) or 40 µM in CM-URA-LEU. In some instances, iron was added back to the media (from a 50 mM FeCl3 in 1 M HCl stock). For example, BPS(2.5) is BPS(0) media with 2.5 µM iron added back, and YPD(200) is YPD media with 200 µM iron added. All chemicals were obtained from Sigma unless otherwise noted. All restriction enzymes, ligase, and Klenow were obtained from New England Biolabs Inc. (Beverly, MA).
Plasmids and GenesThe S. pombe genes fio1+ and fip1+ were identified through BLAST searches of the NR data base using the FET3 and FTR1 amino acid sequences. The fio1+ gene is Swiss Prot accession number Q09920[GenBank], open reading frame c1F7.08 from chromosome I, EBI accession number Z67998[GenBank]. The fip1+ gene is Swiss Prot accession number Q09919[GenBank], open reading frame c1F7.07C from chromosome I, EBI accession number Z67998[GenBank]. The FET3 and the FTR1 sequences from S. cerevisiae have already been described (4, 5). Sequence comparisons were done with the Bestfit program from the GCG sequence analysis package (Genetics Computer Group, Madison, WI). Multicopper oxidase motifs and potential N-linked glycosylation sites were identified using the Motifs program from GCG.
To generate the GALFET3 plasmid, the plasmid pFET3 (3) containing the genomic FET3 region was digested with Eco0109I and ClaI and treated with Klenow to generate blunt ends. This fragment was purified using the Qiaex II gel extraction kit (QIAGEN, Inc., Chatsworth, CA). The yeast galactose expression vector pJV containing the GAL1 promoter and a LEU2 selectable marker (obtained from Dr. Joni Johnston, University of Utah, Salt Lake City, UT) was digested with BamHI, treated with Klenow, incubated with calf intestinal phosphatase, and purified using the Qiaex II gel extraction kit. The vector and the insert were ligated together using standard procedures. Plasmid, which contained insert in the proper orientation (GALFET3), was transformed into the fet3 disruption strain 1397-6A using the standard lithium acetate transformation procedure, and colonies that contained plasmid were selected on leucine minus medium.
To generate the GALFTR1 and GALFIP1 plasmid, the FTR1 and
fip1+ genes were isolated by PCR using primers
that corresponded to the start and stop regions of the putative open
reading frames. The primers FTR1-1
(5-ccccaagcttgccatgcctaacaaagtgtt-3
) and FTR1-2
(5
-gctctagagaaactcccaccctgtgctag-3
) were used by PCR to isolate the
FTR1 gene from S. cerevisiae F113 genomic DNA on an Idaho
Technologies thermal cycler with the following conditions: 94 °C
30 s hot start/94 °C, 0 s/50 °C, 0 s/72 °C,
30 s for 40 cycles. The primers FIP1-1
(5
-accgaagcttaccttctctaccatggcaaa-3
) and FIP1-2
(5
-ctagtctagacaagcatgactactttaatg-3
) were used to isolate the
fip1+ gene from S. pombe (FY254)
genomic DNA using the same conditions as above. The FTR1-1 and FIP1-1
primers contained HindIII sites, and the FTR1-2 and FIP1-2
primers contained XbaI sites. The PCR products were digested
with HindIII and XbaI, purified from a 1%
agarose gel using the Qiaex II gel extraction kit, and ligated into the
pYES2 galactose expression vector from Invitrogen (San Diego, CA)
digested with HindIII and XbaI. The resulting
plasmids (GALFTR1 and GALFIP1) were transformed into 1397-6A, and
colonies that contained plasmid were selected on uracil minus
medium.
To generate the GALFIO1 plasmid, the primers FIO1-1
(5-tcccaagcttttcttcttttctctcttgcg-3
) and FIO1-2
(5
-ggctctagattattttatttcatcttttc-3
) were made corresponding to
the beginning (22 base pairs upstream of the putative ATG) and the end
of the fio1+ open reading frame. The
fio1+ gene was isolated by PCR from S. pombe (FY254) genomic DNA using the conditions reported above.
These primers also contained XbaI and HindIII
restriction sites. The fragment was digested and cloned into the pYES2
vector. The resulting plasmid pYFIO1 was digested with XbaI
and HindIII, treated with Klenow, and the fragment
containing the fio1+ gene was purified and
blunt-end cloned into the pJV vector. This vector was transformed into
1397-6A, and colonies with plasmid were selected on leucine minus
medium.
Iron transport was measured as described previously (1). Briefly, cells were harvested at optical densities between 0.5 and 2.0 and washed 3 times in assay buffer (containing either galactose or glucose depending on the growth medium of the cells). The cells were incubated in 59Fe (NEN, Boston, MA) at the indicated concentration for 10 min at 30 °C. Assays were done in the presence or absence of 1 mM ascorbate. After the incubation, the cells were washed on glass filters and the 59Fe counts were measured. At concentrations of iron below 1.0 µM, only radioactive iron was added to the assay medium. At concentrations above 1.0 µM, radioactive iron was diluted with cold iron, and the specific activity adjusted accordingly. The overlap between diluted iron and straight radioactive iron at concentrations 0.5 and 1.0 did not show any significant deviations.
Northern AnalysisCells were grown in the indicated medium
and harvested at optical densities between 0.5 and 2.0. Total RNA was
isolated by conventional glass bead/phenol-chloroform extraction. RNA,
10 µg, was run on a 1% agarose/formaldehyde-MOPS gel, blotted onto nylon membrane, and UV-crosslinked. The blots were presoaked in Rapid-hyb buffer (Amersham) for 40 min. Probe was made from the restriction fragments and PCR products mentioned above for
FET3, fio1+, and
fip1+. The FTR1 probe was made from a
PCR product corresponding to the final 300 base pairs of the
FTR1 open reading frame using the primers FTR1-2 and
FTR1-3 (5-gtcgtgactgggaaaaccctggcgagcgaatagtacgcaccgat-3
). A
BstBI restriction fragment containing the S. cerevisiae actin gene was gel purified. A 32P-labeled
probe was made from the DNA fragments using the Prime-It II labeling
kit and purified using the NucTrap Probe purification columns from
Stratagene (La Jolla, CA). FET3, FTR1,
fio1+, and fip1+ probes
were hybridized at 65 °C for 3 h with 1 × 106
counts/ml of probe. For actin probe, hybridization was done at 55 °C
for 3 h. The blots were washed 4 × 15 min in 2 × SSC,
0.1% SDS, and exposed to film for 1 (FET3,
fio1+, fip1+,
FTR1, actin with S. cerevisiae RNA) to 3 h
(S. Pombe RNA with actin).
Studies indicate that high affinity iron transport in
S. cerevisiae is regulated at the level of transcription of
FET3 and FTR1 (8). If these two molecules alone
are sufficient for iron transport, then their iron-independent
expression should permit high affinity iron transport. To examine this
possibility, FET3 was placed under the control of a
galactose promoter and transformed into a fet3 disruption
strain. Under low iron conditions, the presence of galactose allowed
the expression of FET3 and reconstituted iron transport
(Fig. 1). Under high iron conditions, little iron transport was seen although Fet3p was expressed to levels approximately equal to that seen in cells induced in low iron medium (data not shown). Stearmann et al. (4) showed that Ftr1p and Fet3p
must be synthesized simultaneously for high affinity iron transport activity. We confirmed this result by placing FTR1 under
galactose control as well. The fet3 disruption strain was
transformed with plasmids expressing FET3 and
FTR1 under the control of the galactose promoter. When these
cells were grown in galactose high iron medium, no significant increase
in iron uptake was observed unless ascorbate was added to the assay to
bypass the need for ferrireductase activity (Fig. 1). In the presence
of ascorbate, cells grown in high iron medium expressing both
FET3 and FTR1 displayed an increased rate of iron
transport compared with cells expressing FET3 alone. When grown in low iron galactose containing medium, cells expressing both
FET3 and FTR1 show even higher levels of ferrous
iron transport. This transport activity probably reflects induction of
both endogenous and galactose driven FTR1 transcription. The
observation that iron-independent expression of FET3 and
FTR1 induces high affinity iron transport activity
demonstrates that the expression of these two proteins is sufficient
for high affinity iron transport activity.
Characterization of S. pombe Ferrous Iron Transport
To
determine the characteristics of iron transport in S. pombe,
we measured iron accumulation in the S. pombe wild-type
strain FY254. S. pombe and S. cerevisiae
wild-type (F113) cells were grown overnight in YPD and inoculated into
iron-free medium. At specified times, cells were harvested and ferrous
transport was assayed (Fig. 2A). Incubation
in low iron medium resulted in an increased rate of iron transport. The
kinetics and magnitude of the induction of S. pombe iron
transport were similar to that of S. cerevisiae.
To define the concentration dependence of S. pombe iron transport, FY254 and F113 cells were inoculated into low or high iron medium, grown for 9.5 h, and assayed for iron transport using different concentrations of iron (Fig. 2B). At low concentrations of iron, S. pombe displays a concentration dependence of iron transport that is similar to S. cerevisiae. The apparent Km for transport in S. pombe (0.2 µM) is nearly the same as the Km for S. cerevisiae (0.15 µM). At higher concentrations of iron, however, the rate of iron transport decreases in S. pombe, whereas it remains high in S. cerevisiae.
Identification of FET3 and FTR1 Homologues in S. pombeData
base analysis revealed genes in S. pombe similar to
FET3 and FTR1. The putative S. pombe
homologue of FET3, fio1+ (ferrous
iron oxidase) shows 38% identity and 60% similarity on the amino acid
level (Fig. 3A). The S. pombe
Fio1p shows hydrophobic regions at the N terminus and near the C
terminus, similar to those found within the Fet3p (5). As in Fet3p,
Fio1p also has two multicopper oxidase motifs and possesses all
canonical ligands necessary for copper binding (3, 12). Fio1p, like
Fet3p, has 13 potential N-linked glycosylation sites, and 8 of these sites are present in the same location as sites within the
Fet3p.
Sequence comparison of the S. pombe Fio1p and Fip1p to the S. cerevisiae Fet3p and Ftr1p. A, sequence comparison between the S. cerevisiae oxidase Fet3p and S. pombe Fio1p. Comparisons between the amino acid sequence of FET3 and fio1+ were done using the Bestfit program from the GCG sequence analysis package. Underlined sequences correspond to potential N-linked glycosylation sites. Sequences in bold correspond to multicopper oxidase motifs. B, Sequence comparison between the S. cerevisiae permease Ftr1p and S. pombe Fip1p. Comparison was done using the amino acid sequence of FTR1 and fip1+. Sequences in bold represent the REGLE domain. Underlined sequences represent potential N-linked glycosylation sites.
The putative FTR1 homologue is fip1+ (ferriferous permease), which is 46% identical and 70% similar to Ftr1p on the amino acid level (Fig. 3B). The S. pombe Fip1p has hydrophobic regions similar to Ftr1p and contains the REGLE domain, which may be important for iron binding (4). Both fip1+ and fio1+ are located on chromosome I and are oriented next to each other separated by a common upstream region, suggesting that they may be controlled by a common promoter.
Northern Analysis of fio1+ and fip1+Both S. cerevisiae genes,
FET3 and FTR1, are regulated by iron deprivation;
transcription is increased during conditions of iron limitation (3, 4,
8). Northern blot analysis on iron-deprived cells revealed that
transcription of the S. pombe genes
fio1+ and fip1+ was
regulated by iron need (Fig. 4). Both
fio1+ and fip1+
transcripts were induced when cells were grown in iron-limited conditions (BPS(0)). Alternatively, when cells were grown in
iron-replete conditions, the level of fio1+ and
fip1+ transcripts was reduced. Though the
S. cerevisiae and S. pombe genes are similar,
cross hybridization between FET3 and
fio1+ or FTR1 and
fip1+ was not detected under the hybridization
conditions used in this study.
Iron Transport of a fio1 Disruption Strain
To determine
whether fio1+ played a role in high affinity
iron transport, a fio1 disruption strain, 4051, was
generated, and iron transport activity of this strain was assayed (Fig.
5). The cells were grown overnight in YPD, inoculated
into iron-limited or iron-replete media, grown for 6 h, and
assayed for iron transport. The fio1 disruption strain grown
in low or high iron conditions showed no iron transport activity.
Within the same experiment, wild-type (FY254) cells showed normal
levels of transport. The inability of 4051 to transport low
concentrations of iron indicates a direct role for the
fio1+ gene in high affinity iron transport.
Examination of the concentration curve for wild-type S. pombe iron transport again suggests that there is a significant
inhibition of the S. pombe high affinity transport system at
high iron concentrations even with incubations as short as 10 min.
Expression of fio1+ and fip1+ in the S. cerevisiae fet3 Disruption Strain
To further define the role of
the S. pombe gene fio1+ in iron
transport, this gene was cloned into a S. cerevisiae
galactose expression vector and transformed into the fet3
disruption strain 1397-6A. The presence of
fio1+ within an S. cerevisiae strain
lacking FET3 did not result in complementation of the iron
transport defect (Fig. 6). This result may indicate that
the S. pombe ferroxidase, while homologous to Fet3p, is not
similar enough to allow assembly with the S. cerevisiae permease, Ftr1p. The S. pombe fip1+ permease
gene under the control of the galactose expression system was
introduced into the fet3 disruption strain containing the galactose-regulated fio1+. These cells were
grown under galactose high and low iron conditions, and iron transport
was measured. Expression of both S. pombe genes resulted in
rates of iron transport above that seen in cells expressing either gene
or vector alone (Fig. 6). This result indicates that expression of both
S. pombe fio1+ and fip1+
genes is sufficient to reconstitute a functional iron transport system.
Genetic studies have shown that both FTR1 and FET3 are required for high affinity iron transport in S. cerevisiae. Other genes are also necessary for iron transport, as they are required for production and localization of active Fet3p. Two genes, CTR1 and CCC2, encode copper transporters that are necessary for copper to be incorporated into Fet3p (6, 7, 13). AFT1, a transcriptional regulator of FET3, FTR1, and the ferrireductases FRE1 and FRE2, is also necessary (8, 14, 15). A dominant mutation of AFT1 has been identified that results in the constitutive transcription of genes involved in high affinity iron transport (8). In this strain, iron transport is elevated even in high iron conditions. These and other results show that transcription alone controls the high affinity iron transport system in S. cerevisiae (1, 8). Normally, under high iron conditions, genes involved in high affinity iron transport are not induced (8). Our results demonstrate the expression of both FET3 and FTR1 independent of iron need results in high affinity ferrous iron transport even when cells are grown in iron-replete medium. Given the caveat that there may be molecules critical for iron transport that are constitutively expressed, this result shows that induction of FET3 and FTR1 are sufficient to effect ferrous iron transport.
Physiological and genetic studies indicate that S. pombe has a high affinity inducible iron transport system that is very similar to S. cerevisiae. The Km for transport and kinetics of induction in low iron conditions are nearly identical in the two species. A difference we observed is that exposure of S. pombe to high iron resulted in a rapid decrease in the rate of iron transport (Fig. 2B and Fig. 5). In S. cerevisiae, however, exposure of cells to high iron produces changes in iron transport over hours not minutes (1). The rapidity of the change in transport rate in S. pombe suggests a post-translational effect. The reduction in transport activity due to high iron is not seen in S. cerevisiae expressing the S. pombe genes (data not shown), suggesting that any regulatory effect is not intrinsic to the transporter proteins. Thus, S. pombe iron transport may be regulated by genes that are not present in S. cerevisiae.
The S. pombe genes fio1+ and fip1+ show amino acid similarity, analogous motifs, and hydrophobic regions corresponding to their S. cerevisiae counterparts. Both fio1+ and fip1+ gene transcripts are also induced when cells are grown in iron-depleted conditions. Disruption of fio1+ in S. pombe results in a loss of high affinity iron transport activity. Expression of both fio1+ and fip1+ in S. cerevisiae mediates high affinity iron transport. Together these results strongly support the conclusion that these genes constitute a high affinity ferrous iron transport system. Expression of the S. pombe gene fio1+ alone in an S. cerevisiae fet3 disruption strain did not result in complementation of the fet3 defect. Although fio1+ and FET3 are homologous in function, Fio1p is not able to interact with Ftr1p and mediate iron transport.
The FET3-FTR1 based system of transmembrane iron transport in S. cerevisiae is complex. This system was unexpected as most other transition metal transport systems require only a transmembrane transporter. For example, in S. cerevisiae, transport of iron by the low affinity transport system (FET4) seems to be simpler, involving the activity of just one gene (2). We have suggested that the oxidase-permease complex is required for high affinity iron transport, rather than a simple transmembrane transporter, because the ferroxidase imbues selectivity on the transport system (5). High affinity iron transport results from a complex set of reactions requiring reduction of iron followed by its subsequent re-oxidation by the cell surface ferroxidase (Fet3p) and transport of Fe (III) by the transmembrane permease (Ftr1p) (3, 4, 5, 10, 14). Identification of an oxidase-permease transport system in the evolutionary distinct fission yeast S. pombe demonstrates that this complex system is not restricted solely to S. cerevisiae and may have a broad distribution in eukaryotes.
The authors wish to acknowledge the assistance of Drs. J. P. Kushner and R. J. Ajioka as well as our colleagues in the Kaplan laboratory for help in preparing this manuscript. We also thank Dr. Susan Forsburg and her lab for supplying the S. pombe strains, vectors, and advice.