Introduction

Iron is essential to nearly all organisms. Cellular iron exists mainly in the form of iron-containing cofactors heme and iron–sulfur clusters. Heme is involved in diverse biological processes, such as transport, cell respiration, drug metabolism, ion channel function and gene regulation (Liu et al. 2020; Fleischhacker et al. 2021). Iron–sulfur clusters are present in proteins involved in ATP production, metabolism, DNA synthesis and repair, gene expression regulation, protein translation, and the antiviral response (Rouault and Maio 2017; Lill and Freibert 2020).

In organisms including fungi (Bairwa et al. 2017, Misslinger et al. 2021), plants (Kobayashi et al. 2019), and bacteria (Bradley et al. 2020), iron is taken up by two basic strategies under iron-deficient conditions. One strategy involves the cell surface Fe3+ reductase and a high-affinity iron-uptake complex composed of a ferroxidase and a Fe2+ permease. Extracellular Fe3+ is first reduced to the more soluble Fe2+ form by a Fe3+ reductase such as Fre1 from Saccharomyces cerevisiae and Frp1 from Schizosaccharomyces pombe. Fe2+ is then re-oxidized by a ferroxidase such as Fet3 from S. cerevisiae and Fio1 from S. pombe, and transported by an iron permease such as Ftr1 from S. cerevisiae and Fip1 from S. pombe. Notably, the S. cerevisiae Fre1 requires heme for its activity because cells defective in heme biosynthesis lack ferrireductase activity (Lesuisse and Labbe 1989) and Fet3 is a multicopper oxidase (Askwith et al. 1994).

The second strategy involves siderophores, which are small Fe3+-specific chelators. Siderophores are secreted to take up extracellular Fe3+. The Fe3+–siderophore complex is then taken up by the cell via specific transport systems. In S. pombe, the cell surface protein Str1 has been found to take up siderophore-bound iron (Pelletier et al. 2003). Unlike S. cerevisiae, which cannot synthesize siderophore and use siderophore synthesized by other microorganisms, S. pombe is able to synthesize the hydroxamate-type siderophore ferrichrome (Schrettl et al. 2004; Mercier and Labbe 2010). The siderophore synthetase Sib1 and the ornithine N5-oxygenase Sib2 are two critical enzymes involved in ferrichrome biosynthesis in S. pombe.

In addition, S. cerevisiae employs the low-affinity iron-uptake system under iron-replete conditions. This system also uses cell surface ferric reductases to convert extracellular Fe3+ into Fe2+. Fe2+ is then transported by the plasma membrane protein Fet4 (Dix et al. 1994). In addition to Fet4, another plasma membrane protein Smf1 from S. cerevisiae is also involved in low-affinity iron uptake (Chen et al. 1999). However, transports involved in the low-affinity iron uptake have not been identified in S. pombe. In addition to the basic iron-uptake strategy, S. pombe also makes use of extracellular heme acquisition to take up iron (Mourer et al. 2015, 2017, 2019). During this process, the cell surface membrane protein Shu1 is required for high-affinity uptake of heme, and heme is initially accumulated in vacuoles and then distributed in the cytoplasm. Besides Shu1, Str3 is involved in low-affinity heme acquisition in S. pombe (Normant et al. 2018).

In this study, we identified the homologs of S. cerevisiae two high-affinity iron transporter (Fre1 and Fet3) and two low-affinity iron transporters (Fet4 and Smf1) in fungi, and analyzed the protein domains of selective homologs. We also examined the effects of deletion of S. pombe homologous genes on the sensitivity of S. pombe cells to iron and other transition metals. Our studies suggested that S. pombe two high-affinity iron transporter Fio1 and Frp1 may be involved in the uptake of transition metals in addition to iron. Our studies also suggested that iron uptake and copper uptake may be coupled in S. pombe. The analysis provides a foundation for future studies aimed at elucidating the detailed coupling mechanism of iron and copper uptake in S. pombe.

Materials and methods

Strains, plasmids and media

Schizosaccharomyces pombe strains used in this study are described in Table 1. The PCR primers used in this study are described in Table S1. Gene deletions were made by the one-step gene replacement using a kanamycin resistance gene (kanMX6) or hygromycin resistance gene (hphMX6) as a selectable marker (Hentges et al. 2005). Constructs for the deletion mutants were generated as described below. Briefly, the 5′- and 3′-flanking regions of the target gene were amplified by PCR using the S. pombe genomic DNA as a template, and cloned into the Sal I/Bgl II and Sac I/Spe I sites of the plasmid pAF6a-kanMX6 (Bahler et al. 1998) or pAF6a-hphMX6 (Wang et al. 2017), respectively. To generate single deletion mutants, the deletion cassettes were amplified from corresponding deletion constructs by PCR and transformed into the wild-type strain yHL6381 using the lithium acetate method (Forsburg and Rhind 2006). The deletion cassettes were transformed into the Δfio1 strain to generate double deletion mutants. Transformants were selected on YES plates containing 100 μg/mL of geneticin (G418) or 100 μg/mL of hygromycin. All deletion mutants were verified by PCR.

Table 1 Strains used in the present study
Full size table

Bacterial cells were grown in liquid or solid LB medium (1% tryptone, 0.5% yeast extract, 1% sodium chloride, pH 7.2) without or with 100 μg/mL ampicillin to select for plasmids at 37 °C. S. pombe cells were routinely grown at 30 °C in liquid or solid YES medium containing 0.5% yeast extract, 3% glucose and 225 mg/L adenine, leucine, histidine, and uracil.

Standard media and protocols for genetic manipulation of fission yeast were used as described (Forsburg and Rhind 2006).

Metal sensitivity assays

Schizosaccharomyces pombe cells were grown overnight at 30 °C in YES medium. Cells were diluted to an OD600 of 0.2 in fresh YES medium, and grown to early exponential phase. Cells were collected and adjusted to the same OD600 of 3.0. Serial tenfold dilutions were prepared and 2 μL of each dilution was spotted onto YES plates supplemented with 25 µM bathophenanthroline disulfonic acid (BPS), 2.75 mM Fe2(SO4)3 (Fe3+), 800 µM CuSO4 (Cu2+), 800 µM ZnSO4 (Zn2+), 800 µM MnSO4 (Mn2+), or 80 µM CdCl2 (Cd2+). The plates were incubated at 30 °C for 4–7 days and photographed. Initially, different concentrations of metal ions were tested based on the previous studies (see below). CuSO4 was tested at 0.5, 0.8 and 1.5 mM (Li and Kaplan 1998). ZnSO4 was tested at 0.5, 0.8, 2, and 3.5 mM (Li and Kaplan 1998). MnSO4 was tested at 0.5, 0.8, 2 and 3.5 mM (Li and Kaplan 1998). CdCl2 was tested at 80 and 100 µM (Kennedy et al. 2008). Fe2(SO4)3 was used at 2.75 mM based on our previous study (Su et al. 2017). BPS was used at a concentration of 25 µM based on a previous study (Encinar del Dedo et al. 2015).

Bioinformatics analyses

Homologs of S. cerevisiae Fet3, Fre1, Fet4 and Smf1 were identified through BLAST (Basic local alignment search tool) searches of the NCBI non-redundant protein sequences database (https://blast.ncbi.nlm.nih.gov/Blast.cgi) with default parameters. The candidate proteins were then analyzed for the presence of the conserved domain using the Conserved Domain Database (CDD, https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) (Lu et al. 2020) and Simple Modular Architecture Research Tool (SMART, http://smart.embl-heidelberg.de/) (Letunic et al. 2021). Protein sequences were aligned using the MUSCLE algorithm included in the Molecular Evolutionary Genetics Analysis (MEGA) software (version 10.1.8) (Tamura et al. 2021). Sequence logos were generated using WebLogo 3 (http://weblogo.threeplusone.com/create.cgi).

Results

Schizosaccharomyces pombe high-affinity iron transporter Fio1 contains the cupredoxin-like domains

We first performed BLAST searches against protein databases using the S. cerevisiae multicopper ferroxidase Fet3 as a query. Fet3 homologs were identified in Ascomycota, including Pezizomycetes (792 species), Saccharomycotina (1021 species) and Taphrinomycotina (39 species), and Basidiomycota (735 species), as well as in the basal clades of fungi, including Zygomycota (375 species) and Chytridiomycota (79 species) (data not shown). This result suggested that Fet3 homologs are universally present in fungi. Analysis of the protein sequences using the CDD and SMART databases predicted that like S. cerevisiae Fet3, S. pombe Fio1 (a homolog of S. cerevisiae Fet3) and other representative fungal Fet3 homologs contain three cupredoxin-like domains linked by external interdomain (Fig. 1). The cupredoxin domain typically adopts a Greek-key, β-barrel structure, comprising two β-sheets formed by 7 or 8 β strands. This structure was initially found in small copper-containing electron-transfer proteins in bacteria, fungi, and plants termed blue-copper proteins or cupredoxins, and later found in multicopper oxidase (Sedlak et al. 2008).

Fig. 1
figure 1

Saccharomyces cerevisiae Fet3 and its homologs contain three cupredoxin-like domains. a Domain organization of representative Fet3 proteins from S. pombe, S. cerevisiae, Aspergillus fumigates, Candida albicans, Coprinopsis cinerea, Fusarium coicis, Saccharata proteae, and Lophium mytilinum. Domains were identified by CDD and SMART analyses. The cupredoxin-like domains are indicated in the blue boxes. b Multiple sequence alignment of the cupredoxin-like domain sequences of representative Fet3 proteins. The alignment was generated using MEGA_X_10.1.8. The copper-binding residues are marked with an asterisk above the alignment. Amino acid residues are highlighted according to the biochemical properties of the amino acid residues. The sequence logo was generated using WebLogo 3

Full size image

Multiple sequence alignment revealed that S. cerevisiae Fet3 and its homologs contain ten absolutely conserved histidine residues and one absolutely conserved cysteine residue located in two cupredoxin domains, suggesting that these histidine and cysteine residues may be involved in copper binding in Fio1. Indeed, these residues have been shown to be important for the binding of four coppers. Fet3 contains three different copper centers involved in the redox reaction. It remains to be determined that Fio1 is a multicopper oxidase.

Schizosaccharomyces pombe Frp1 and Frp2 contain the ferric reductase domain

A BLAST search using the S. cerevisiae Fre1 protein sequence revealed that Fre1 homologs are ubiquitously present in fungi. S. pombe contains two Fre1 homologs (Frp1 and Frp2). The two proteins share 47% identity and 63% similarity in the amino acid sequence over the entire length. Frp1 is localized to the cell surface, whereas Frp1 resides in the endoplasmic reticulum (ER) (Matsuyama et al. 2006). SMART database search predicted six transmembrane domains in both Frp1 and Frp2 proteins. CDD and SMART analyses revealed that both Frp1 and Frp2 and other representative Fre1 homologs contain a conserved ferric reductase domain (Fig. 2) (Zhang et al. 2013). This domain includes three canonical conserved heme-coordinating histidines (Fig. 2). Multiple sequence alignment of Fre1 and selected Fre1 homologs revealed two additional absolutely conserved histidine residues, which are likely involved in catalysis and/or heme-binding (Fig. S1).

Fig. 2
figure 2

Saccharomyces cerevisiae Fre1 and its homologs contain the ferric reductase and transmembrane domains. a Domain organization of S. cerevisiae Fre1 and its homologs from S. pombe (Frp1 and Frp2), A. fumigates, C. albicans, C. cinerea, F. coicis, S. proteae, and L. mytilinum. The ferric reductase and transmembrane domains identified by CDD and SMART are indicated in blue and red boxes, respectively. b Multiple sequence alignment of the ferric reductase domain sequences of representative Fre1 proteins. The conserved heme-coordinating histidines are indicated with asterisks. Amino acid residues are highlighted according to the biochemical properties of the amino acid residues. The alignment and the sequence logo were generated as described above

Full size image

Schizosaccharomyces pombe putative low iron affinity transporters Fet4 and Pdt1 are integral transmembrane proteins

We went on to identify homologs of two S. cerevisiae low-affinity iron transporters Fet4 and Smf1 using BLAST searches. Fet4 and Smf1 homologs are widely present in fungi. We selected the representative candidates of Fet4 and Smf1 for further analysis. CDD and SMART database searches showed that S. cerevisiae Fet4 and representative Fet4 homologs contain an iron-permease domain consisting of 9–11 transmembrane regions (Fig. S2). An examination of the amino acid sequence of the iron-permease domain from S. cerevisiae Fet4 and its homologs did not reveal any obvious candidate metal-binding ligands in the hydrophilic regions of the proteins. Such ligands have been found in S. cerevisiae Fet3 and its homologs (see above). The exact mechanism of how this iron-permease domain controls iron uptake into the cell remains to be determined.

CDD analyses revealed that S. cerevisiae Smf1 and its homologs, including S. pombe Pdt1, belong to the NRAMP metal ion transporter family, which is characterized by a conserved hydrophobic core of 10–12 transmembrane regions (Bozzi and Gaudet 2021). SMART analysis revealed the presence of 10 conserved transmembrane regions in the proteins (Fig. S3). Multiple sequence alignment revealed absolutely conserved aspartic acid and glutamic acid residues in the NRAMP domain, which are likely to be involved in catalysis and metal binding (Fig. S3).

Effect of deletion of iron mutants on the growth under iron-replete and iron-depleted conditions

To examine whether S. pombe fio1, frp1, frp2, fet4 and pdt1 are involved in iron uptake, we constructed deletion mutants for each of these genes by homologous recombination and analyzed the growth of these deletion mutants under iron-replete and iron-depleted conditions. To create iron-depleted conditions, 25 µM of iron chelator bathophenanthroline disulfonate (BPS) was added to YES media. For high-iron conditions, Fe2(SO4)3 was added to YES media for a final concentration of 2.75 mM. The BPS and all metal ion concentrations used in the assays were optimized. ∆fio1 cells, and to a much lesser extent, ∆frp1 cells, exhibited a growth defect on iron-depleted media; whereas ∆frp1 cells, and to a much lesser extent, ∆frp2 cells showed a growth defect on iron-replete conditions (Fig. 3a). In contrast, ∆fet4 and ∆pdt1 cells did not exhibit a growth defect under iron-depleted and iron-replete conditions (Fig. 3a). These results suggest that Fio1 is essential for iron uptake under iron-deficient conditions and that additional proteins may be involved in ferric reduction under iron-deficient conditions in S. pombe. These results also suggest that Frp1 and Frp2 may be involved in iron uptake under iron-replete conditions.

Fig. 3
figure 3

The sensitivity of wild-type, Δfio1, Δfrp1, Δfrp2, Δfet4 and Δpdt1 cells to iron, copper, zinc, manganese, and cadmium. a The effect of iron on the growth of wild-type, Δfio1, Δfrp1, Δfrp2, Δfet4 and Δpdt1 cells. Cells were grown in YES medium to exponential phase and adjusted to an OD600 of 3.0. Tenfold serial dilutions were prepared, and 2.5 μL aliquots of each dilution were spotted on YES plates, and YES plates containing 25 µM BPS to generate the iron-depleted growth condition or 2.75 mM Fe2(SO4)3 to generate the iron-replete growth condition. b The sensitivity of wild-type, Δfio1, Δfrp1, Δfrp2, Δfet4 and Δpdt1 cells to copper, zinc, manganese, and cadmium. Cells were prepared as described above, and were spotted on YES plates, and YES plates containing 800 µM CuSO4 (Cu2+), 800 µM MnSO4 (Mn2+), 800 µM ZnSO4 (Zn2+), or 80 µM CdCl2 (Cd2+)

Full size image

Fio1 and Frp1 are likely to be involved in the uptake of transition metals other than iron

Because iron metal transporters may be involved in the transport of metal ions other than iron, we examined the sensitivity of Δfio1, Δfrp1, Δfrp2, Δfet4 and Δpdt1 cells on transition metals Cu2+, Zn2+, Mn2+, and Cd2+. The concentrations of transition metals used in the assay were optimized. We also included Δpap1 cells as a control, because pap1 encodes an oxidative stress-responsive basic leucine zipper (bZIP) transcription factor required for cell survival under metal ion stress. We found that only Δfio1 cells were sensitive to Cu2+. In addition, Δfio1 and Δfrp1 cells were sensitive to Zn2+, Mn2+. In contrast, Δfrp2, Δfet4 and Δpdt1 did not exhibit sensitivity to all metal ion tested. Δpap1 cells were only modestly sensitive Cd2+ (Fig. 3b), suggesting that the addition of Cd2+ activated Pap1, whereas the addition other transition metals did not. These results suggest that Fio1 and Frp1 may be involved in the uptake of transition metals in addition to iron.

Addition of iron can suppress the sensitivity of Δfio1 and Δfrp1 to transition metals

Since Fio1 and Frp1 appear to be involved in the uptake of transition metals in addition to iron, we tested whether the addition of iron affected the growth of wild-type strain and Δfio1, Δfrp1, Δfrp2, Δfet4 and Δpdt1 mutants on media supplemented with transition metal Cu2+, Zn2+, Mn2+, or Cd2+. We found that the addition of iron could suppress the sensitivity of Δfio1 cells to Cu2+, Zn2+, Mn2+, and Cd2+. In addition, the addition of iron could suppress the sensitivity of Δfrp1 cells to Zn2+, Mn2+, and Cd2+ (Fig. 4).

Fig. 4
figure 4

The effect of iron additions on the sensitivity of wild-type, Δfio1, Δfrp1, Δfrp2, Δfet4 and Δpdt1 cells to transition metals. Cells were grown in YES medium to exponential phase, adjusted to an OD600 of 3.0. Tenfold serial dilutions were prepared, and 2.5 μL aliquots of each dilution were spotted on YES plates containing designated concentrations of transition metals with or without indicated concentrations of Fe2(SO4)3

Full size image

Deletion of frp1 can rescue the sensitivity of Δfio1 cells to copper

We genetically tested whether S. pombe Fio1 functions together with Frp1, Frp2 and Fet4 in the uptake of transition metals. To do so, we generated double mutants of Δfio1Δfrp1, Δfio1Δfrp2 and Δfio1Δfet4. We also generated Δfrp1Δfet4 as a control. The results showed that deletion of frp1 but not frp2 and fet4 could rescue the sensitivity of Δfio1 cells on Cu2+. In addition, deletion of frp1, frp2 and fet4 could not rescue the sensitivity of Δfio1 cells to Zn2+ and Mn2+ (Fig. 5). These results suggested that S. pombe Fio1 and Frp1 function together in copper uptake.

Fig. 5
figure 5

Deletion of frp1, but not frp2 or fet4, can rescue the sensitivity of Δfio1 cells to copper. Cells were grown in YES medium to exponential phase, and adjusted to an OD600 of 3.0. Tenfold serial dilutions were prepared, and 2.5 μL aliquots of each dilution were spotted on YES plates, and plates supplemented with transition metals at the indicated concentrations

Full size image

Discussion

In this study, we investigated the role of S. pombe Fio1, Frp1, Frp2, Fet4 and Pdt1 in the uptake of iron and other transition metals. Our iron sensitivity assays suggested that ferroxidase Fio1 and, to a much lesser extent, ferric reductase Frp1 are involved in high-affinity iron uptake, which is required for cell survival under iron-deficient conditions, consistent with the previous finding (Kwok et al. 2006). In addition, our results suggest that Frp1 is likely involved in low-affinity iron uptake. It remains to be determined whether Fet4 and Pdt1, like their S. cerevisiae counterparts, are involved in low-affinity iron uptake.

We tested the sensitivity of Δfio1, Δfrp1, Δfrp2, Δfet4 and Δpdt1 deletion mutants to transition metals such as copper, zinc, manganese, and cadmium. The growth of Δfio1 cells was inhibited at 0.8 mM CuSO4 (Fig. 3). The growth of Δfio1 cells was inhibited even at a lower CuSO4 concentration (0.5 mM) (Fig. S4). A similar result was obtained with a higher concentration of CuSO4 (1.5 mM) (Fig. S4). The growth of Δfio1 and Δfrp1 cells was impaired at 0.8 mM MnSO4 and 0.8 mM ZnSO4. Similar results were obtained with 0.5 and 2 mM MnSO4, and 0.5 mM ZnSO4 (Fig. S4). At high concentrations of MnSO4 (3.5 mM) and ZnSO4 (2 and 3.5 mM), the growth of both wild-type and mutant cells was impaired (Fig. S4). Δfrp1 cells were more sensitive to CdCl2 than wild-type cells at concentrations of 80 and 100 μM (Fig. 3 and Fig. S4). One possible explanation for how deletion of fio1 or frp1 increases sensitivity to transition metals is that defective iron uptake caused by deletion of fio1 or frp1 may increase the uptake of other transition metals, resulting in the elevated levels of transition metals, which are toxic to the cells. These results suggested that like in S. cerevisiae, in S. pombe iron uptake may be coupled to other transition metal uptake. Interestingly, our results revealed that the sensitivity of Δfio1 and Δfrp1 cells to other transition metals could be rescued by iron addition. A similar observation has been made in S. cerevisiae (Li and Kaplan 1998). It is likely that iron addition to Δfio1 and Δfrp1 cells may lead to an increased cellular iron level and a decreased uptake of potentially toxic transition metals.

Our results revealed that the sensitivity of Δfio1 cells to copper can also be rescued by deletion of frp1. This observation may be explained as follows. The ferric reductase Frp1 may also be involved in the reductive uptake of copper in S. pombe. In the copper uptake, Frp1 may produce toxic Cu1+ that may be re-oxidized to Cu2+ by the putative multicopper oxidase Fio1 and the Fio1-generated Cu2+ is then transported into the cell by a permease. Thus, it is likely that failure of re-oxidization of Frp1-produced Cu1+ to Cu2+ by Fio1 results in the sensitivity of Δfio1 cells to copper. In S. cerevisiae, deletion of fre1 suppresses the copper sensitivity of Δfet3 cells (Shi et al. 2003). It is likely that a similar mechanism for coupling iron uptake to copper uptake may operate in S. cerevisiae. Further work is required to validate the proposed coupling mechanism.

We also observed that the sensitivity of Δfio1 cells to Cu2+ cannot be rescued by deletion of frp2. Frp2 is homologous to Frp1, and the two proteins share 47% amino acid sequence identity and 63% similarity. One possible explanation for this is that although the two proteins may perform similar functions, the two proteins have distinct subcellular localization in S. pombe. Like Fio1, Frp1 is distributed on the cell surface, whereas Frp2 is localized to the ER (Matsuyama et al. 2006). It seemed likely that the ER-resident Frp2, unlike the cell surface Frp1, is involved in iron uptake into the ER.