The Role of the FRE Family of Plasma Membrane Reductases in the Uptake of Siderophore-Iron in Saccharomyces cerevisiae *

Saccharomyces cerevisiae takes up siderophore-bound iron through two distinct systems, one that requires siderophore transporters of the ARN family and one that requires the high affinity ferrous iron transporter on the plasma membrane. Uptake through the plasma membrane ferrous iron transporter requires that the iron first must dissociate from the siderophore and undergo reduction to the ferrous form. FRE1 and FRE2 encode cell surface metalloreductases that are required for reduction and uptake of free ferric iron. The yeast genome contains five additional FRE1 and FRE2 homologues, four of which are regulated by iron and the major iron-dependent transcription factor, Aft1p, but whose function remains unknown. Fre3p was required for the reduction and uptake of ferrioxamine B-iron and for growth on ferrioxamine B, ferrichrome, triacetylfusarinine C, and rhodotorulic acid in the absence of Fre1p and Fre2p. By indirect immunofluorescence, Fre3p was expressed on the plasma membrane in a pattern similar to that of Fet3p, a component of the high affinity ferrous transporter. Enterobactin, a catecholate siderophore, was not a

Virtually every organism on earth requires iron as an essential nutrient. Although iron is the second most abundant metal in the crust of the earth, the bioavailability of iron can be extremely low. This poor bioavailability occurs because iron is rapidly oxidized in an aerobic environment to the ferric form (Fe(III)), 1 which is poorly soluble in water and forms precipitates of oxyhydroxides. Microorganisms have the capacity to scavenge iron from insoluble precipitates by secreting and taking up siderophores, low molecular weight compounds that bind to Fe(III) with very high affinity and specificity. Siderophores are synthesized and secreted in the iron-free form, which then binds and solubilizes Fe(III) in the extracellular environment. The Fe(III)-siderophore complex is then recognized and selectively taken up by specific transport mechanisms. Many microorganisms synthesize one or a few types of siderophores, yet have the capacity to take up iron from a variety of siderophores secreted by other species of bacteria and fungi (1). Budding and fission yeast appear to be an exception; they neither synthesize nor secrete these compounds (2,3). Saccharomyces cerevisiae can, however, recognize and take up iron from a variety of structurally distinct siderophores (4 -10).
S. cerevisiae has two genetically separable systems for the uptake of siderophore-bound iron. One system depends on a family of homologous transporters of the major facilitator superfamily that is expressed as part of the AFT1 regulon and are termed ARN1, ARN2 (also TAF1), ARN3 (also SIT1), and ARN4 (also ENB1) (6 -11). These transporters are expressed in intracellular vesicles. The individual ARN transporters exhibit specificity for different siderophores of the hydroxamate and catecholate classes; however, some siderophores, such as rhodotorulic acid, are not substrates of the ARN transporters (9). A second system of uptake for siderophore-bound iron depends on the high affinity ferrous iron (Fe(II)) transport complex, which is encoded by FET3 and FTR1 and is located on the plasma membrane (12)(13)(14)(15). A low affinity Fe(II) transporter encoded by FET4 is also expressed on the plasma membrane (16). For the siderophore-bound Fe(III) to become a substrate for the Fe(II) transporter, the iron must be both reduced and dissociated from the siderophore. This is accomplished in a single step by the activity of plasma membrane reductase systems, which contain flavocytochromes and have the capacity to reduce siderophore-bound iron (4,5,17). FRE1 and FRE2 encode plasma membrane metalloreductases that can reduce oxidized forms of both iron and copper (18 -23). Strains deleted for FRE1 exhibit only 10% of the Fe(III)-citrate reductase activity that is inducible in wild-type strains. Deletion of both FRE1 and FRE2 results in cells that are completely lacking Fe(III)-citrate reductase activity and fail to grow on iron-poor media. The completed sequence of the S. cerevisiae genome revealed the presence of five additional genes with striking similarity to FRE1 and especially to FRE2. Four of these (FRE3, FRE4, FRE5, and FRE6) are greater than 35% identical to FRE2 and are regulated at the transcriptional level by Aft1p (24). The fifth homologue (FRE7) is regulated by exogenous copper ions through the Mac1p transcription factor. The functions of these new FRE family members have not been identified.
We have investigated the role of the FRE family of genes in the uptake of siderophore-bound iron. Although the FRE genes appeared to have no role in the ARN-dependent uptake of siderophores, they were required for the uptake of siderophorebound iron through the high affinity Fe(II) transport system. Although FRE1 and FRE2 encoded the majority of siderophore reductase activity, Fre3p could specifically facilitate reduction and uptake of iron bound to the trihydroxamate siderophores ferrioxamine B (FOB), ferrichrome (FC), and triacetylfusarinine C (TAFC) and to the dihydroxamate rhodotorulic acid (RA). Fre3p was expressed on the plasma membrane in a pattern consistent with its role in iron uptake through the plasma membrane Fe(II) transport system. Uptake of iron bound to the catecholate siderophore enterobactin (ENT) also occurred through the Fe(II) transport system and required either Fre1p or Fre2p. Expression of Fre4p was sufficient to facilitate the utilization of RA-bound iron when the siderophore was present in higher concentration.
Siderophores-Iron-free forms of FC, RA, and the mesylate salt of FOB (Desferal) were purchased from Sigma. Ferric and desferric forms of TAFC and ENT were prepared as described (9,29). Ferric RA, FC, and FOB were prepared by incubating the desferri-siderophores with an equimolar amount of FeCl 3 in 50 mM sodium citrate, pH 6.5, and 5% glucose.
Surface Reductase Assay-Cells were grown to mid-log phase in YPD medium or defined iron media containing 10 M ferrous ammonium sulfate, and the assay was performed as described (18) with the following modifications. Ferrozine was substituted for bathophenanthroline disulfonate (BPS) as the Fe(II) acceptor, and 1 mM Fe(III)-FC, -FOB, -TAFC, and -RA were the reductase substrates. Absorbance of the cell-free assay mixture was measured at 562 nm.
Plate Assay and Uptake Assay-For the plate assay, modified synthetic complete media were used in which copper and iron were omitted, and 1 M copper sulfate and 100 M BPS were added. Trace amounts of iron are supplied by agar (Difco). Plate assays for low concentrations of FC, FOB, TAFC, ENT, and RA were performed as described (10) using 10 M desferri-FC, -FOB and -TAFC, 5 and 10 M ferric-ENT, and 15 M ferric-RA (a concentration with iron-chelating capacity equivalent to 10 M FC, FOB, TAFC, and ENT). Plate assays for high concentrations of FOB and RA were performed using 100 M of the desferri-siderophore. Because of ligand exchange between the BPS and the siderophores, the exact concentration of the ferric-siderophore complex in the plate assay is not known. Uptake assays were performed as described (30) in Multiscreen Filtration plates (Millipore) using 1 M each of 55 Fe(III) and desferri-FOB and-ENT.
Immunofluorescence-Strains YPH499 pFRE3-HA and YPH499 FET3-HA FTR1-Myc were grown to mid-log phase in defined iron media containing 5 M ferrous ammonium sulfate and 1 mM ferrozine to induce the expression of Fre3p. Cells were washed and prepared for immunofluorescence microscopy as described (15). Primary antibody was affinity-purified HA.11 (Babco) at 1:300, and secondary antibody was Cy3-conjugated polyclonal anti-mouse IgG from donkey (Jackson ImmunoResearch) at 1:500. Cells were visualized with a Zeiss 63X/ 1.4NA objective. Images were acquired with a cooled CCD camera (Cooke) using IP Labs software (Scanalytics).

Reduction of Ferric Siderophores by FRE Reductases-FOB-
bound iron can be taken up at the plasma membrane by a reductive mechanism that requires heme (5). The metalloreductase encoded by FRE1 is a plasma membrane heme protein of the cytochrome b-type, and strains deleted for FRE1 exhibit low Fe(III)-citrate reductase activity (17)(18)(19). We tested whether other ferric-siderophores of the hydroxamate class were substrates for cell surface reductases (Fig. 1). In addition to FOB, both the trihydroxamates FC and TAFC and the dihydroxamate RA were substrates for plasma membrane reductases. RA-bound Fe(III), which is only taken up through the Fe(II) transport system, displayed the highest rate of reduction, whereas Fe(III) bound to FC, which is taken up by both FIG. 1. Capacity of the FRE gene products to reduce Fe(III)siderophores in FRE؉ and fre؊ deletion strains. After culture in rich media or defined iron media containing 10 M ferrous iron (inset), cells of the indicated genotype in the exponential phase of growth were collected, washed, and assayed for Fe(III)-siderophore reductase activity in 1 mM Fe(III)-ferrichrome, -ferrioxamine B, -fusarinine, and -rhodotorulic acid with ferrozine as the Fe(II) acceptor. Absorbance at 562 nm was measured. The experiments were repeated twice, and data from a representative experiment are shown. Inset, the fre1,2,3⌬ and the fre1,2⌬ strain were transformed with either a high copy number plasmid bearing the FRE3 gene (pFRE3) or the empty parent plasmid (pRS424) and the Fe(III)-ferrioxamine B reductase activity measured as described. Samples were prepared in triplicate, and the experiment was repeated three times. Error bars indicate the S.D.
the Fe(II) transport system and the Arn1p and Arn3p transporters, displayed the lowest rate of reduction. FRE1 and FRE2 encoded 90 -98% of the siderophore reductase activity, but there were measurable differences in the siderophore reductase activities of a fre1⌬ fre2⌬ strain and a fre1⌬ fre2⌬ strain that also carried a deletion of FRE3 (Fig. 1, inset). Deletion of FRE3 in the fre1⌬ fre2⌬ strain resulted in a 73% decrease in the reduction of FOB. This loss of FOB reductase activity was completely restored when the fre1,2,3⌬ strain was transformed with a plasmid expressing Fre3p, indicating that Fre3p can facilitate the reduction of FOB-iron. These data did not indicate whether the very low level of residual reductase activity associated with expression of Fre3p, Fre4p, Fre5p, and Fre6p could significantly affect iron uptake and utilization by the cell. For this reason, we tested the capacity of FRE-deleted strains to take up Fe(II)-FOB and to grow when FOB was the iron source.
The Role of FRE3 in Ferrioxamine B-mediated Iron Uptake and Growth-FOB-bound iron can be transported into the cell by both an Arn3p-dependent system and a Fet3p-dependent system that also involves an Fe(III) to Fe(II) reduction step. We tested the role of FRE1, FRE2, and FRE3 in the uptake of FOB-iron in strains both expressing Arn3p and deleted for Arn3p ( Fig. 2A). When cells were grown in rich media, deletion of FRE3 alone resulted in an increased uptake of FOB-iron when compared with the congenic parent strain. This effect was consistently reproducible and suggested that deletion of FRE3 resulted in changes in the intracellular iron pool. Deletion of FRE1 and FRE2 resulted in dramatic decreases in FOB-iron uptake, both in an ARNϩ and an arn3⌬ background, indicating that Fre1p and Fre2p reductase activities contribute to FOB-iron uptake. Deletion of FRE3 in addition to FRE1 and FRE2 resulted in a small, additional decrease in FOB-iron uptake in both the ARNϩ and the arn3⌬ backgrounds, so that there was no detectable FOB-iron uptake in the arn3⌬ fre1⌬ fre2⌬ fre3⌬ strain. To determine whether Fre3p contributed significantly to FOB-iron uptake, Fre3p was overexpressed from a high copy-number vector in the arn3⌬ fre1⌬ fre2⌬ fre3⌬ strain (Fig. 2B). Fre3p overexpression resulted in a slight increase in uptake when FOB-iron was present at 1 M and a larger increase in uptake when FOB-iron was present at 10 M. These data indicated that Fre3p could facilitate the reductive uptake of FOB-iron.
To clarify further the role of FRE1, FRE2, and FRE3 in the utilization of FOB-iron, we examined the growth of FRE-deleted strains on iron-limited media containing FOB as an iron source (Fig. 2C). We used strains in which ARN-dependent FOB uptake was inactivated by deletion of ARN3; therefore, iron uptake was dependent on Fet3p and the high affinity Fe(II) uptake system. An arn3⌬ fre1⌬ fre2⌬ strain did not grow in the absence of FOB, but grew as well as the FREϩ parent strain in the presence of FOB, indicating that the reductase activities encoded by FRE1 and FRE2 are not absolutely required for utilization of FOB-iron. An arn3⌬ fre1⌬ fre2⌬ fre3⌬ strain, however, failed to grow in the presence of FOB, confirming that Fre3p can facilitate the uptake and utilization of FOB-iron. This result also indicated that endogenous levels of expression of Fre4p, Fre5p, Fre6p, and Fre7p could not support the reductive uptake of FOB-iron. Deletion of FRE3 alone had no effect on growth on FOB, and transformation of the arn3⌬ fre1⌬ fre2⌬ fre3⌬ strain with a low copy number plasmid carrying either the FRE1 or FRE3 gene completely restored growth on FOB (data not shown). Fre1p, Fre2p, and Fre3p are not required for uptake of FOB through the ARN-dependent system, as an ARNϩ fre1⌬ fre2⌬ fre3⌬ strain grew well on FOB (Fig. 2C).
The Role of Fre3p in the Uptake of Ferrichrome, Triacetylfusarinine C, and Rhodotorulic Acid-Because Fre3p could facilitate the reductive uptake of FOB-iron, we questioned whether structurally related siderophores were also substrates for Fre3p. The iron-binding sites of FC and TAFC are similar in structure to that of FOB in that they each consist of three bidentate hydroxamate ligands that bind Fe(III) in a six-coordinate complex. We examined the capacity of the FRE homologues to facilitate growth when FC and TAFC in low concentration (10 M) were provided as iron sources (Fig. 3, A and B). Again, these experiments were performed using strains in which the ARN-dependent siderophore uptake systems were inactivated by deletion of ARN1 and ARN3 to test FC and by deletion of ARN2 to test TAFC. Deletion of either FRE3 alone or FRE1 and FRE2 together had no effect on FC-mediated growth (Fig. 3A). Deletion of FRE1, FRE2, and FRE3 in combination, however, resulted in a failure to grow on FC, indicating that Fre3p can facilitate reductive iron uptake from both FC and FOB. As was the case for FOB, deletion of FRE1, -2, and -3 in an ARNϩ background did not impede FC-mediated growth, further supporting the role of the FRE genes in siderophore-iron uptake through the Fe(II) transport system but not through the ARN system. Similar results were obtained when a strain deleted for FRE1, -2, and -3 was tested for TAFC-mediated growth (Fig. 3B), indicating that Fre3p can facilitate reductive iron uptake from each of the three trihydroxamate siderophores tested.
RA-mediated iron uptake in S. cerevisiae significantly differs from FOB-, FC-, and TAFC-mediated uptake in that RA is not a substrate for the ARN transporters and RA-iron uptake occurs exclusively through the Fe(II) transport system (9). Also, RA is a dihydroxamate siderophore that coordinates Fe(III) in a RA 3 Fe 2 complex (31). Despite these differences, Fre3p also facilitated RA-iron-mediated growth (Fig. 3C), as the fre1⌬ fre2⌬FRE3ϩ strain grew as well as the congenic parent strain on ferric-RA, whereas the fre1⌬ fre2⌬ fre3⌬ strain did not grow.
Localization of Fre3p to the Plasma Membrane-Both components of the high affinity Fe(II) transport system, Fet3p and Ftr1p, are expressed on the plasma membrane (10,(13)(14)(15). In contrast, the Arn1p, Arn3p, and Arn4p transporters are largely confined to intracellular vesicles that comigrate on density gradients with Pep12p, a protein of the late endosome (9, 10) (data not shown). To determine the cellular localization of Fre3p, we constructed a strain in which the chromosomal copy of FRE3 carries a triple copy of the HA epitope at the carboxyl terminus. We confirmed that the epitope-tagged Fre3p was functional by expressing the HA-tagged Fet3p from a plasmid in a FRE3-deleted strain and observing restoration of FOB-iron uptake to wild-type levels. We grew the Fre3-HAp strain and a strain expressing an HA-tagged version of Fet3p in media containing limiting amounts of iron (to induce Fre3-HAp expression) and performed indirect immunofluorescence (Fig. 4). Fre3p was detected at the periphery of the cell in a pattern that was similar to that of Fet3p, indicating that both Fre3p and Fet3p were expressed on the plasma membrane. Fre3p was not detected in intracellular vesicles. This localization to the plasma membrane was consistent with a role for Fre3p in reductive iron uptake at the plasma membrane through the high affinity Fe(II) system. The absence of detectable Fre3p in intracellular vesicles is also consistent with genetic data indicating that Fre3p is not required for ARN-dependent siderophore uptake.
Reductive Uptake of Fe(III)-Enterobactin-ENT is a siderophore of the catecholate class that is synthesized and secreted by species of Gram-negative bacteria, such as E. coli. S. cerevisiae was recently reported to take up ENT through the ARN4 transporter (7), but it is not known whether ENT-bound iron can be taken up via a reductive mechanism that requires the high affinity Fe(II) transport system. We examined the uptake of ENT-bound iron in strains bearing a deletion of ARN4 alone and deletions of all four ARN genes, both in strains with intact high affinity Fe(II) transport (FET3ϩ) and in strains lacking high affinity Fe(II) transport (fet3⌬, Fig. 5A). Deletion of ARN4 in both FET3ϩ and fet3⌬ backgrounds resulted in decreases in ENT-iron uptake to undetectable levels, confirming that uptake of ENT-iron is facilitated by Arn4p. Uptake of ENT-iron was increased over 10-fold in a fet3⌬ strain, and this was associated with a large increase in the level of expression of the Arn proteins, especially Arn4p, in the fet3⌬ strain (data not shown). This increase likely reflects the relative iron deprivation and subsequent Aft1p activation that results from the loss of high affinity Fe(II) uptake in fet3⌬ strains. These results suggested that the majority of ENT-iron uptake occurs through Arn4p, but whether ENT-iron could also be taken up through a reductive mechanism remained unclear.
To evaluate ENT-iron uptake via the high affinity Fe(II) system, we tested the capacity of ENT to stimulate growth of strains on iron-limited media (Fig. 5B). Strains individually deleted for ARN4 or FET3 grew as well as the wild-type parent strain in the presence of ENT. Deletion of both ARN4 and FET3 resulted in a strain that did not grow in the presence of ENT, indicating that ENT is a substrate for both the ARN-dependent and FET3-dependent systems of iron uptake. To determine the role of the FRE genes in FET3-dependent ENTiron uptake, we tested the capacity of ENT to stimulate growth of FRE-deleted strains that were also deleted for ARN4 (Fig.  5C). In this experiment, the arn4⌬ FREϩ and the arn4⌬ fre1⌬ strains grew well on iron-limited media in the presence of ENT and very slowly in the absence of ENT. In contrast, the arn4⌬ fre1⌬ fre2⌬ and the arn4⌬ fre1⌬ fre2⌬ fre3⌬ strains exhibited virtually no growth in the presence or absence of ENT. Expression of FRE1 from a low copy number plasmid restored ENTmediated growth to the arn4⌬ fre1⌬ fre2⌬ strain (data not shown), indicating that either Fre1p or Fre2p is also required for the FET3-dependent utilization of ENT-iron.
The Role of Fre4p in Rhodotorulic Acid Iron Utilization-Fre4p, Fre5p, and Fre6p did not appear to have a role in the reductive uptake of low concentrations of siderophore-bound iron, because endogenous levels of expression of these proteins in fre1⌬ fre2⌬ fre3⌬ strains did not provide sufficient activity to result in growth on low concentrations of siderophore. This apparent lack of function could occur because of the following: 1) FRE4, FRE5, and FRE6 are not expressed as functional enzymes on the plasma membrane; 2) the siderophores we tested were not substrates for Fre4p, Fre5p, and Fre6p; or 3) the tested siderophores were substrates but exhibited lower affinity for Fre4p, Fre5p, and Fre6p. We tested the last hypothesis by examining the growth of FRE-deleted stains on a higher concentration of RA and FOB. In Fig. 6, each of the six ironregulated FRE genes was serially deleted, and the resulting strains were tested for the capacity to grow in the presence of 100 M RA and FOB. Deletion of FRE1, FRE2, and FRE3 had no effect on the capacity of a high concentration of RA to stimulate growth on iron-limited plates (Fig. 6A). Deletion of FRE1, FRE2, FRE3, and FRE4, however, resulted in a strain that grew very slowly on iron-limited media in the presence or absence of 100 M RA. In contrast, a high concentration of FOB did not stimulate growth of the arn3⌬ fre1⌬ fre2⌬ fre3⌬ strain. These data indicate that Fre4p could specifically facilitate the reductive uptake of RA-iron, but not FOB-iron, through the Fe(II) transport system. DISCUSSION Uptake of siderophore-bound iron through the high affinity Fe(II) transport system requires the activity of plasma membrane reductases that are encoded by the FRE family of genes.
Here we have shown that Fre1p and Fre2p can facilitate the reduction of ferric-siderophores of the trihydroxamate and di-hydroxamate classes. This finding was consistent with previous observations that the plasma membrane reductase system of budding yeast exhibits an extremely broad range of substrate specificity that includes both ferric and nonferric electron acceptors (32). We also demonstrated that Fre3p could facilitate the reduction of FOB and that Fre3p and Fre4p could facilitate the reductive uptake and utilization of siderophoreiron, presumably by acting as ferric-siderophore reductases. Although Fre4p could facilitate utilization of RA-iron, we were not able to detect significant Fre4p-dependent RA-iron reductase or uptake activities. The most likely explanation for these observations is that S. cerevisiae can grow in the face of extremely low rates of iron uptake, and it is less likely that Fre4p facilitates iron uptake through an activity other than that of a reductase. Fre3p and Fre4p exhibit the most similarity to Fre2p with 72 and 56% identity, respectively, at the amino acid level (24). Fre5p and Fre6p exhibit lesser degrees of similarity to Fre2p (38 and 35%, respectively), and the function of these gene products remains unknown. Although the siderophore reductase activity exhibited by Fre3p is 40-fold lower than that of Fre1p and Fre2p, FRE3 and FRE4 mRNA transcripts are present in quantities roughly comparable to those of FRE1 and FRE2 (24). This observation suggests that the differences in activity are due less to differences in the rate of transcription and more to differences in the intrinsic properties of the proteins. We observed that overexpression of Fre3p did not result in increased Fre3p-specific reductase activity when compared with the activity associated with endogenous levels of expression. This observation suggests the possibility that the plasma membrane reductase system has additional components that may be present in limiting quantities.
Although Fre1p, Fre2p, Fre3p, and Fre4p could facilitate the reductive uptake of siderophore-iron, they were not functionally interchangeable. Fre3p and Fre4p did not facilitate growth on low concentrations of ferric-ENT, but Fre3p could facilitate growth on low concentrations of di-and trihydroxamates. Fre4p did not facilitate growth in the presence of low concentrations of catecholate siderophore or hydroxamate siderophores nor on high concentrations of the trihydroxamate FOB. Yet Fre4p did facilitate growth on high concentrations of the dihydroxamate RA. These results suggest that the FRE reductases differ both in their specificities and affinities for different siderophores. Further evidence of differing specificities is seen in the case of the FRE3 deletion strain. Deletion of FRE3 in cells grown in rich media leads to an increased rate of uptake of FOB-iron, which might occur if Fre3p facilitated uptake of a significant amount of iron from a specific (but as yet unknown) source. Thus, in the absence of Fre3p, cells would respond to the loss of iron uptake from this specific source by up-regulating remaining systems of iron uptake. An alternative explanation for the increase in iron uptake in the absence of Fre3p is that Fre3p is acting as both a reductase and a sensor of iron. Although externally directed environmental iron sensors have not been reported in budding yeast, a two-component Fe(III) sensor has recently been identified in Salmonella enterica (33).
The differences in specificity exhibited by the FRE reductases may be based on differences in the structure of the iron coordination site in the siderophores. Alternatively, differences in specificity may be accounted for by the reduction potentials of the siderophores. In Table I, the capacities of siderophores to stimulate growth of FRE-deleted strains and the reduction potentials (31) of these siderophores are presented. ENT had both the most negative reduction potential and the most stringent reductase requirements, stimulating Arn4p-independent growth only when Fre1p or Fre2p was expressed. Although the data presented here suggest that Fre1p or Fre2p can reduce ENT-iron, ENT, with a reduction potential of Ϫ750 mV, is thought to be beyond the range of physiological reductants (31). This may not be the case in an in vivo setting, however, as a positive shift in the reduction potential of the siderophore may occur when the pH is lowered, the concentration of the reduced species is very low, or the hydrophobicity of the microenvironment is altered. In contrast to ENT, RA exhibits both the least negative reduction potential and the least stringent reductase requirements, serving as a substrate for Fre1p, Fre2p, Fre3p, and, at higher concentrations, Fre4p. Whether a particular ferric-siderophore can be reduced by one of the FRE reductases may depend more on whether the reductase can generate sufficient reduction potential than on the capacity of the reductase to "recognize" a structural motif.
Reductive uptake of iron at the plasma membrane requires the function of both an Fe(III)-reductase complex (containing a FRE gene product and possibly other components (34)) and an Fe(II) transport complex (containing Fet3p and Ftr1p). Fet3p and Ftr1p are physically associated, and their intracellular association is required for efficient expression on the plasma membrane (13)(14)(15). Although deletion of the FRE genes did not impede the expression of the Fet3p-Ftr1p complex on the plasma membrane, 2 the reductase and transport complexes may be associated as part of a plasma membrane-based iron uptake "machine." TABLE I Comparison of growth stimulation capacity and reduction potential of siderophores Data from Figs. 2, 3, 5, and 6 are presented. The capacity of the siderophore at a concentration of 10 M to stimulate growth of strains with the indicated FRE genotype was scored (ϩ, wild type growth; Ϫ, minimal growth). Growth on plates containing 100 M RA is reported in parentheses. Reduction potentials (31) are listed in mV. Each reduction potential was measured at pH 7 in dilute aqueous solution versus a standard hydrogen electrode. Variations in activity coefficients, due to differences in buffer type or ionic strength, are generally assumed to be Ͻ10 mV.