Download - Iron use for haeme synthesis is under control of the yeast frataxin homologue (Yfh1

Iron use for haeme synthesis is under controlof the yeast frataxin homologue (Yfh1)

Emmanuel Lesuisse1,*, Renata Santos1, Berthold F. Matzanke2, Simon A. B. Knight3,

Jean-Michel Camadro1 and Andrew Dancis3

1Laboratoire d’Ingenierie des Proteines et Controle Metabolique, Departement de Biologie des Genomes, Institut

Jacques Monod, Unite Mixte de Recherche 7592 CNRS-Universites Paris 6 and 7, 2 place Jussieu, F-75251 Paris

cedex 05, France, 2University of Luebeck, Isotopes Laboratory TNF, Ratzeburger Allee 160, D-23538 Luebeck,

Germany and 3Department of Medicine, Division of Hematology-Oncology, University of Pennsylvania, Philadelphia,

PA 19104, USA

Received December 20, 2002; Accepted February 7, 2003

The YFH1 gene is the yeast homologue of the human FRDA gene, which encodes the frataxin protein.Saccharomyces cerevisiae cells lacking the YFH1 gene showed very low cytochrome content. In Dyfh1strains, the level of ferrochelatase (Hem15p) was very low, as a result of transcriptional repression of HEM15.However, the low amount of Hem15p was not the cause of haeme deficiency in Dyfh1 cells. Ferrochelatase, amitochondrial protein, able to mediate insertion of iron or zinc into the porphyrin precursor, made primarilythe zinc protoporphyrin product. Zinc protoporphyrin instead of haeme accumulated during growth of Dyfh1mutant cells and, furthermore, preferential formation of zinc protoporphyrin was observed in real time. Themethod for these studies involved direct presentation of porphyrin to mitochondria and to ferrochelatase ofpermeabilized cells with intact architecture, thereby specifically testing the iron delivery portion of the haemebiosynthetic pathway. The studies showed that Dyfh1mutant cells are defective in iron use by ferrochelatase.Mossbauer spectroscopic analysis showed that iron was present as amorphous nano-particles of ferricphosphate in Dyfh1 mitochondria, which could explain the unavailability of iron for haeme synthesis. A highfrequency of suppressor mutations was observed, and the phenotype of such mutants was characterized byrestoration of haeme synthesis in the absence of Yfh1p. Suppressor strains showed a normal cytochromecontent, normal respiration, but remained defective in Fe–S proteins and still accumulated iron intomitochondria although to a lesser extent. Yfh1p and Hem15p were shown to interact in vitro by Biacorestudies. Our results suggest that Yfh1 mediates iron use by ferrochelatase.

INTRODUCTION

The YFH1 gene is the yeast homologue of the human FRDAgene, which encodes the frataxin protein. Mutations of FRDAassociated with decreased frataxin expression are responsiblefor Friedreich’s ataxia, the most common autosomal-recessiveneurodegenerative disease of Caucasians (1,2). Both genescode for mitochondrial proteins that are involved in ironhomeostasis and cellular respiration (3–8), but their preciseroles are unknown. Cardiac tissues from patients withFriedreich’s ataxia exhibit iron deposition, deficiencies in manyiron–sulphur cluster enzymes and reduced mitochondrial DNA(8,9). In addition, fibroblasts from these patients showhypersensitivity to oxidative stress that can be rescued by

treatment with iron chelators (10). A link to haeme biosynthesishas not been uncovered, and blood, bone marrow and red celldevelopment appear to be normal in patient with frataxindeficits. Frataxin is downregulated during erythroid develop-ment, suggesting that this protein is not involved in the high-volume iron trafficking that accompanies red cell production inthe bone marrow (11). However, activities of haeme enzymes inother tissues of Friedreich’s ataxia patients have not beenassessed, leaving open the possibility that tissue-specific haemedeficiencies may exist.

Yeast cells lacking Yfh1p mirror many of the phenotypesobserved in disease tissues from patients with Friedreich’sataxia. These cells have defective respiration (3,5,12–14),unstable mitochondrial DNA and hypersensitivity to oxidative

*To whom correspondence should be addressed. Tel: þ33 144276356; Fax: þ33 144275716; Email: [email protected]

Human Molecular Genetics, 2003, Vol. 12, No. 8 879–889DOI: 10.1093/hmg/ddg096

Human Molecular Genetics, Vol. 12, No. 8 # Oxford University Press 2003; all rights reserved

by guest on May 31, 2014

http://hmg.oxfordjournals.org/

Dow

nloaded from


stress (3–5). The assembly of Fe–S centres is impaired andcytochrome concentrations are low (4,8). Iron uptake is muchgreater than in wild-type cells, with most of the iron beingfound in the mitochondria (3,12). The involvement of Yfh1p inthe assembly of Fe–S centres has been described in severalstudies (8,15–18), but no work has been devoted yet toinvestigate the possible specific involvement of Yfh1p in thesynthesis of haeme, a major iron cofactor synthesized into themitochondria. An impediment to understanding the function ofYfh1p or frataxin has been the complex nature of the cellularphenotypes resulting from depletion or loss of function. Herewe reexamine the role of Yfh1p in iron homeostasis withspecial emphasis on haeme synthesis. We describe a switchfrom haeme synthesis to zinc protoporphyrin synthesis thatoccurs in absence of Yfh1p. A highly sensitive fluorimetricmethod is used to demonstrate this switch. Previous studieshave not noted effects of Yfh1 on haeme formation in yeast,and this may be due to the high frequency of suppressormutations that mask this phenotype. Here we describe thecharacteristics of such suppressor mutants, and the effects onhaeme formation in the absence of Yfh1p.

RESULTS

Cytochrome deficiency in strains lacking Yfh1p

Yeast cells lacking the YFH1 gene grew slowly (or not at all,depending on the strain) on non-fermentable carbon sources,consistent with a defect in mitochondrial function. The coloniesappeared depigmented and measurement of total cellular haemerevealed global haeme deficiency in the mutant cells (0.01 nmolhaeme/mg dry weight) compared with the wild-type (0.2 nmolhaeme/mg dry weight). Low-temperature spectra of Dyfh1mutant whole cells revealed a virtual absence of b, c and(a þ a3) cytochrome signals (Fig. 1). In contrast, a signal fromzinc protoporphyrin was observed in the mutant cells, and thissignal was further enhanced by zinc supplementation of thegrowth media. This observation is significant because bio-synthesis of Fe-PPIX (haeme) and Zn-PPIX both requireporphyrin precursor and ferrochelatase activity, differing onlyin the final metal insertion step. The prevalence of Zn-PPIXsynthesis rather than haeme synthesis in Dyfh1 cells did notresult from increased zinc accumulation in Dyfh1 cellscompared with wild-type cells. Actually, the total zinc contentof the cells was lower in Dyfh1 cells than in wild-type cells(402 mg zinc per g Dyfh1 cell paste and 457 mg zinc per g wild-type cell paste). A detailed study of zinc metabolism in Dyfh1cells will be published elsewhere.

A general problem that has impeded characterization ofYfh1p function is the variability and instability of thephenotypes of the deletion strains. In part this is due to atendancy to lose functional mtDNA, thereby becoming rhominus or rho zero. A mutator phenotype characterized byincreased frequency of secondary nuclear mutations has alsobeen associated with Yfh1p loss of function (19). Therefore, inorder to distinguish primary effects from secondary changesensuing from lack of YFH1, we created a strain in which thesole copy of YFH1 was placed under the control of a galactoseinducible promoter. In this strain, regulated expression of Yfh1

allowed correlation of phenotypes with different Yfh1pexpression levels. If this strain was grown in raffinose, a non-inducing carbon source, for 24 h, Yfh1p was undetectable byimmunoblotting of isolated mitochondria, and the cellsexhibited phenotypes similar to the Dyfh1 strain. However,under these conditions, rho minus conversion or secondarygenetic changes were not observed (not shown). As wasobserved for the deletion strain, cytochromes were undetectablein the low temperature spectra, whereas Zn-PPIX was clearlydiscerned as a 580 nm absorbance peak (Fig. 2). Cytochrome cwas an abundant haeme protein of the intermembrane spacewas undetectable by blotting in these cells. When the cellswere exposed to galactose, the promoter was rapidly inducedand Yfh1p expression reached a maximum within 2 h.Cytochromes in general and cytochrome c in particular wererecovered, although the time course of recovery seemeddelayed with respect to the recovery of Yfh1p levels. Yfh1pwas completely restored at the 2 h time point, whereascytochromes recovery lagged behind (Fig. 2). The criticalYfh1p function involved in haeme formation presumablyoccurs during this time interval.

Loss of function of Yfh1 affects the final step ofhaeme synthesis

The final step in haeme biosynthesis involves iron insertioninto PPIX and is catalysed by ferrochelatase, the product of theHEM15 gene. The Hem15 protein, localized to the mitochon-drial inner membrane, was markedly decreased in Dyfh1mitochondria (Fig. 3A). HEM15 mRNA was also lower in theDyfh1 mutant than in wild-type cells (Fig. 3B), as also reportedby others (20). Other key enzymes of the haeme pathway wereunchanged in yfh1 mutant cells: the amounts of Hem1p

Figure 1. Prevalence of Zn-PPIX synthesis rather than haeme synthesis inDyfh1 cells. Low-temperature spectra of whole cells of the strain S150-2B(curve 1) and S150-2BDyfh1 (curve 2). The cells were grown for 15 h inYPR medium and then transferred to a YPGal medium containing 0.5 mM

ZnSO4 for a further 15 h. The arrow indicates the maximum absorption peakof Zn-PPIX.

880 Human Molecular Genetics, 2003, Vol. 12, No. 8



Dow

nloaded from


(d-aminolevulinate synthase) and Hem13p (coproporphyrino-gen oxidase) were the same as in the wild-type (Fig. 3A), andthe activity of Hem14p (protoporphyrinogen oxidase) wasnormal (see below). These data suggested that the defect inhaeme synthesis in yfh1 mutant cells might be due to low levelof ferrochelatase protein. We therefore sought to correct theHem15p deficiency by using a multicopy plasmid to increaseexpression of the gene in the Dyfh1 strain. Surprisingly,the Dyfh1 cells transformed with YEp351-HEM15 grew evenmore slowly than the untransformed cells, forming tiny colonieson agar plates one week after transformation (not shown). Thetotal haeme and cytochrome contents of the transformedmutant cells remained much lower than the wild-type cells (notshown). Thus, the cytochrome defect of Dyfh1 cells does notresult from a lack of Hem15p. Moreover, while cytochromeswere almost undetectable in Dyfh1 cells, a peak of Zn-PPIXwas clearly apparent in these cells (see below). Zinc is analternative substrate of ferrochelatase (21). Thus, althoughHem15p was expressed at very low levels in the Dyfh1 mutant,the protein was still produced and probably functional in themutant (see below).

An unresolved question was why Hem15 levels were low inthe Dyfh1 strain. HEM15 transcription was found to beregulated in an iron-dependent manner, without dependenceon Aft1p or Aft2p, the iron regulatory transcription factors(22,23). As shown in Figure 3B, iron addition to the growthmedia was correlated with a 2–3-fold increase in transcriptabundance whether or not Aft1 or Aft2 was present. In theDyfh1 strain the transcript was virtually undetectable. The

molecular mediators of this iron dependent regulation remainto be determined. It is generally admitted that Aft-dependentgenes are upregulated in Dyfh1 cells because cytosolic iron islow in these cells, most of the iron being sequestered in themitochondrial compartment (14). In the case of HEM15, anAft-independent gene, downregulation of transcription wasprobably again mediated by alterations in cytoplasmic ironlevels, but the regulatory pathway appears to be a novel one.

Iron unavailability for haeme synthesis in Dyfh1 cells

Studies of in vitro haeme synthesis are often difficult tointerpret, since ferrochelatase has a high affinity for ferrousiron, and in vivo iron availability to ferrochelatase probablydepends on crucial factors related to iron compartmentalizationor availability of an electron donor. Therefore, addition ofexogenous ferrous ions and protoporphyrin IX (PPIX) toisolated mitochondria will result in haeme synthesis in vitro,even if no haeme synthesis occurred in vivo.

The level of ferrochelatase protein in Dyfh1 cells was lowerthan in wild-type cells, although the residual protein levelvaried depending on the yeast genetic background fromroughly 10% (X498-1A) to 25% (S150-2BDyfh1) of normal.This residual ferrochelatase protein was active as shown bythe presence of Zn-PPIX in the mutants. Furthermore,the residual ferrochelatase in Dyfh1 mutants was able tomediate haeme formation (as measured by the pyridinehaemochromogen method) when iron as ferrous ascorbate (or

Figure 2. Low-temperature spectra of whole cells of the strain GAL-Yfh1 var-ious times after induction of YFH1 transcription. Cells of the GAL-Yfh1 strainwere grown for 24 h on YPR, diluted 20-fold into fresh YPR medium andgrown for another 16 h. At this time (t¼ 0), part of the cells were harvestedand galactose (2%) was added to the culture. Cell aliquots were harvested2 and 5 h after galactose addition. Low-temperature spectra of whole cells wererecorded (left panel). Yfh1p and Cyc1p levels were estimated by blotting of themitochondrial proteins (right panel). The figure shows one experiment out ofthree independent experiments done.

Figure 3. Blotting of proteins of the haeme biosynthesis pathway in various iso-genic yeast strains (A), and iron-dependent and Aft-independent transcriptionof HEM15 (B). The strains used were all isogenic to S150-2B. (A). Whole cellextracts (Hem1, Hem13) or mitochondrial extracts (Hem15) were prepared afterovernight growth of the cells on YPD þ 15mg/ml hemin. Proteins were sepa-rated by SDS–PAGE before being blotted and revealed with the specificHem1, Hem13 and Hem15 antibodies. (B). Total RNA was probed withHEM15 and ACT1 by northern blotting. Strains were grown in minimum mediawithout iron or supplemented with 50 mM Fe(III)–citrate.

Human Molecular Genetics, 2003, Vol. 12, No. 8 881



Dow

nloaded from


ferric citrate þ NADH) and PPIX were added to isolatedmitochondrial membranes. By contrast, haeme in Dyfh1 cells(X498-1A or S150-2B) was virtually undetectable, leading usto conclude that an additional defect in iron or porphyrinavailability to ferrochelatase must exist in these cells. We thendeveloped an in vitro assay to measure endogenous ironavailability to ferrochelatase, using permeabilized whole cells(Fig. 4A) or intact mitochondria (Fig. 4B). We usedprotoporphyrinogen instead of PPIX as the substrate ofreaction, and no exogenous metals, so that only endogenousiron or zinc could be incorporated into PPIX to form haeme orZn-PPIX. Protoporphyrinogen is the substrate of protoporphyr-inogen oxidase, an inter-membrane space enzyme, whichconverts protoporphyrinogen into PPIX. Protoporphyrinogenand haeme are both non-fluorescent, while PPIX and Zn-PPIXare both highly fluorescent. The use of protoporphyrinogen assubstrate allowed monitoring with high sensitivity the rate ofPPIX and Zn-PPIX synthesis. In the presence of a metalchelator such as EDTA or 8-hydroxyquinoline, haeme and Zn-PPIX formation from PPIX was inhibited, and we measured therate of PPIX formation from protoporphyrinogen, i.e. proto-porphyrinogen oxidase activity. This activity was comparablein both wild-type and yfh1 mutant cells (Fig. 4A, left panel),showing that the porphyrin substrate for ferrochelatase was notthe limiting factor for haeme synthesis. When no chelator wasadded, part of the PPIX synthesized could be used byferrochelatase to form haeme or Zn-PPIX with endogenousmetals. The rate of haeme synthesis can then be calculated as:total PPIX (measured in the presence of a chelator) minus (Zn-PPIX þ PPIX) (measured without chelator). In the absence of achelator, much more PPIX accumulated in Dyfh1 cells than inthe wild-type (Fig. 4A, middle panel), suggesting that PPIXwas being produced but not utilized for haeme synthesis in theDyfh1 cells. Finally, Zn-PPIX was formed more rapidly inDyfh1 cells than in wild-type cells, consistent with theunavailability of iron as a substrate for ferrochelatase(Fig. 4A, right panel). Similar experiments were done withintact isolated mitochondria from the GAL-Yfh1 strain(galactose-dependent expression of YFH1) repressed forYFH1 expression, and induced for 2 h and 5 h with galactose(Fig. 4B). Similar activity of protoporphyrinogen oxidase wasobserved with all mitochondria samples (Fig. 4B, upper leftpanel). However, when endogenous metals were not chelatedby EDTA, the rate of formation of PPIX from protoporphyr-inogen decreased as YFH1 was induced (Fig. 4B, upper rightpanel), meaning that an increasing part of the PPIX producedwas used to form Zn-PPIX and haeme as YFH1 becameinduced. The rate of Zn-PPIX formation with endogenous zincwas comparable in all the mitochondria samples (Fig. 4B,bottom left panel). In contrast, the availability of endogenousiron to form haeme was very low in mitochondria withrepressed Yfh1, and increased after galactose induction,especially after 5 h induction (Fig. 4B, bottom right panel).This result fits well with the data of Figure 2, showinginduction of cytochrome synthesis in cells of the same strainafter 2 and 5 h induction of YFH1 by galactose. It is interestingto note that induction of Yfh1 expression in Yfh1-deficientcells did not result in immediate restoration of haeme synthesisand cytochrome production: while the level of Yfh1p wascomparable to wild-type levels after 2 h induction by galactose,

Figure 4. Synthesis of PPIX, Zn-PPIX and haeme in permeabilized whole cells(A) or isolated mitochondria (B) from protoporphyrinogen and endogenousmetals. (A) PPIX (a) and Zn-PPIX (b) production by permeabilized wholewild-type (1) or Dyfh1 (2) cells. The reaction was initiated by adding 2mM pro-toporphyrinogen to the cell suspension (OD 10). Formation of PPIX (lexc

410 nm, lem 632 nm) and of Zn-PPIX (lexc 420 nm, lem 587 nm) was followedfluorometrically. (a) Protoporphyrinogen oxidase activity; PPIX synthesis wasmonitored after chelation of all endogenous metals (iron and zinc) by adding10 mM 8-hydroxyquinoline to the permeabilized cell suspension. (b) Synthesisof PPIX in the presence of endogenous metals. (c) Synthesis of Zn-PPIX inthe presence of endogenous metals. (B) Cells of the strain GAL-Yfh1 weregrown on YPR. YFH1 transcription was either repressed (t¼ 0; squares) orinduced for 2 h (t¼ 2 h; triangles) or for 5 h (t¼ 5 h; circles) by galactose, asfor Figure 2. Mitochondria were isolated from the cells and incubated at100mg/ml in isotonic buffer (0.5 M sorbitol, 0.1 M TRIS, pH 7.4) with (upper leftpanel) or without (the other three panels) EDTA (1 mM). The reaction wasinitiated by adding 2 mM protoporphyrinogen to the suspension of mitochondria.Formation of PPIX (lexc 410 nm, lem 632 nm) and of Zn-PPIX (lexc 420 nm,lem 587 nm) was followed simultaneously (upper panels and bottom left panel).Haeme formation (bottom right panel) was calculated as total PPIX (measured inthe presence of EDTA) minus (Zn-PPIX þ PPIX) (measured without EDTA).The figure shows one experiment out of three independent experiments done.




Dow

nloaded from


the synthesis of haeme (Fig. 4B) and of cytochromes (Fig. 2)was maximum only 5 h after induction.

Iron precipitation in an inorganic form in Dyfh1mitochondria

The observation that iron was unavailable for haeme synthesisin Dyfh1 cells raised a question of why this should be so. Thephysical state of the iron was examined in mutant mitochondriausing Mossbauer spectroscopy. Figure 5 shows Mossbauerspectra of mitochondria purified from a wild-type (YPH499)recorded at 4.3 K in a small perpendicular field of 20 mT. NoMossbauer signal was visible, indicating an iron concentrationtherein of lower than 300 mM and a lack of iron accumulation.In contrast, the mitochondria of the isogenic Dyfh1 mutant(X498-1A) displayed a well-resolved quadrupole doublet at4.3 K (spectrum not shown) exhibiting an isomer shiftd¼ 0.53(4) mm s�1, a quadrupole splitting DEQ¼ 0.63(1)mm s�1 and a line width G¼ 0.57(1) mm s�1 (numbers inbrackets correspond to calculated error of last digit). No ferrousiron was observed. These Mossbauer parameters are typical ofa high-spin ferric iron bound to oxygen/nitrogen in anoctahedral arrangement. Very similar parameters were foundfor various bacterioferritins at this temperature (24–27).Mitochondria of a Dyfh1 suppressor strain isolated fromX498-1A (sup4þ, see Fig. 5) exhibited a doublet with almostthe same Mossbauer parameters (d¼ 0.52 mm s�1,DEQ¼ 0.67 mm s�1, G¼ 0.53 mm s�1; not shown). However,the degree of ferric ion accumulation per gram mitochondriawas approximately one-quarter of that found in the originalDyfh1 strain.

A G-value of 0.57 mm s�1 indicates a line-width broadeningwhich can be associated with relaxation or superparamagneticphenomena. Indeed, further broadening of the Mossbauer lines[d¼ 0.53(4) mm s�1, DEQ¼ 0.64(1) mm s�1, G¼ 0.73(1)mm s�1] occurred at 1.9 K (Fig. 5). Moreover, the formationof a second unstructured component (42% of absorption area)was observed. In contrast to what was found in bacterioferri-tins, no indication for a distinct magnetic hyperfine field or anarrow ranged field distribution was visible. This and thefeatures of a high field spectrum (7 T, not shown) are consistentneither with a superparamagnetic transition as observed inbacterioferritins, nor with a magnetic transition of antiferro-magnetically m-oxo-coupled systems. The featureless broad-ening is best explained by a broad distribution of individualhyperfine fields originating from many magnetically non-equivalent ferric ions. Thus, our data are consistent with thepresence of small and very amorphous nano-particles of iron inDyfh1 mitochondria. Various attempts to visualize theseparticles on PAGE failed. The material remained in the wellsof the gel, as seen by Fridovich staining (not shown). Therewere only very small amounts, if any, of protein associated withthese particles (0.1 mg protein/mg iron), which could representnon-specific adsorption. Phosphate and iron determinationresulted in a Fe/P ratio of 1/2.9 (8). We conclude that iron wasessentially present in Dyfh1 mitochondria as nano-particles offerric phosphate. In fact, an EXAFS analysis (not shown)supported the structural model of ferric phosphate as the mainiron compound of Dyfh1 mitochondria. A complete analysis of

EPR-, Mossbauer spectroscopic and EXAFS data from wholecells and from mitochondria will be published elsewhere.

Suppressor mutations frequently mask the cytochromedefect of Dyfh1 cells

Some authors described normal cytochrome production inDyfh1 cells (28). According to this observation, the cyto-chrome and respiratory defects reported by others for Dyfh1mutants (3) could depend on a particular genetic background,on the growth conditions, or could result from rho minusconversion of the cells. Our results, however, do not supportthis hypothesis. We constructed a Dyfh1 shuffle strain wherethe yfh1 deletion was covered by a shuffle plasmid bearing awild-type copy of YFH1. Cells of this strain formed isolated,depigmented colonies when plated on YPD þ cycloheximide,but did not grow on YPG þ cycloheximide plates (not shown).When Dyfh1 cells from a YPD plate were inoculated in liquidmedium with glycerol as the carbon source, growth wasdelayed by a lag of 1–3 days (not shown). The cells harvestedafter 5 days on glycerol medium showed a normal cytochromespectrum, unlike cells grown on raffinose as the carbon source,which were completely depigmented (Fig. 6). Cells from theglycerol culture did not recover the original phenotype ofDyfh1 (lack of cytochrome) when re-inoculated on a raffinose-based medium (Fig. 6). This result indicates that some inheri-table change(s) occurred to the cells during their growth onglycerol. Indeed, Dyfh1 cells accumulated suppressor muta-tions with high frequency, which was easily observed on agarplate. When a mat of Dyfh1 cells was plated onto a YPG plate,numerous colonies grew on a background of non-growing(or poorly growing) cells (Fig. 7). The same observation was

Figure 5. Mossbauer spectra of isolated mitochondria from a wild type strain ofS. cerevisiae [YPH499; (A)] and an isogenic Dyfh1 mutant [X498-1A; (B)].The Mossbauer spectra were recorded as described in Methods, at 4.3 K forthe wild-type (A) and at 1.9 K for the Dyfh1 mutant (B).




Dow

nloaded from


made with Dyfh1 cells from different genetic backgrounds,including YPH499, S150-2B, CM3260 and W303 (notshown). We analysed several suppressor colonies of Dyfh1from various genetic backgrounds. Most of the time, thesuppressor strains exhibited the same phenotype as presentedin Fig. 8. Cells recovered a normal cytochrome content(Fig. 8A) and normal activity of haeme-containing enzymessuch as catalase (not shown). The activity of enzymescontaining an iron–sulphur cluster remained low (Fig. 8B).Ferrochelatase and cytochrome c levels were increased com-pared to the original Dyfh1 mutant (Fig. 8C). Respiratoryactivity was similar to that of the wild-type (Fig. 8D). Cell ironaccumulation decreased compared with the original Dyfh1mutant but was still higher than in the wild-type (Fig. 8E), andiron still accumulated in the mitochondria, although to a lesserextent than in the original Dyfh1 mutant (not shown).Resistance of the cells to oxidative stress was increased(Fig. 8F). The suppressor phenotype of glycerol-growing cellsresulted from nuclear mutation(s).

We crossed Dyfh1 cells showing the suppressor phenotypewith an original Dyfh1 mutant of the same background withthe opposite mating type. Features of the diploid wereintermediate between the suppressor strain and the non-suppressed deletion strain, indicating semi-dominance of thesuppressor mutation (Fig. 8C). Following sporulation andtetrad dissection, the suppressor phenotype was recovered inthe tetrads, showing a 2:2 segregation of the suppressorcharacteristics (Fig. 8C). Such nuclear (semi-dominant) sup-pressor mutations occurred with high frequency in Dyfh1 cellssubmitted to the selection pressure of a non-fermentablecarbon source or oxidative stress. The high rate of newsuppressor mutations has prevented us from identifying thesuppressor gene by complementation. A genomic library was

constructed from a Dyfh1 suppressor strain and used totransform an original Dyfh1 strain. Transformants wereselected on a copper-rich medium allowing growth ofsuppressors but not of original Dyfh1 cells. All the coloniesanalysed were new suppressor strains (not shown).

Thus, the presence of normal cytochrome concentration insome Dyfh1 strains (28) may result from suppressormutations rather than from adaptation of the cells toparticular growth conditions. Conversely, the lack ofcytochromes in Dyfh1 was not a consequence of rho minusconversion of the cells, since induction of Yfh1 by galactosein the GAL-Yfh1 strain rapidly induced synthesis of all thecytochromes, with concomitant disappearance of Zn-PPIX, asshown above (Fig. 2). We conclude that Yfh1 is required fornormal cytochrome synthesis in Dyfh1 cells, independentfrom the background and from the tendency of cells to loosemitDNA.

Hem15p and Yfh1p interact in vitro

Our biochemical experiments suggested that the final step ofhaeme biosynthesis required both Hem15p (ferrochelatase) andYfh1p. We therefore looked for a direct interaction betweenferrochelatase and Yfh1p using a real-time biomolecularinteraction analyser, Biacore 2000, based on plasmon surfaceresonance measurements. Ferrochelatase (100 or 500 RU) wasimmobilized on research-grade CM5 sensor chips (Biacore)using a standard amine coupling procedure. Control experi-ments were run using bovine serum albumin (100 or 500 RU)and blank flow cells (activated carboxyl groups reacted withexcess ethanolamine). A specific, high-affinity interactionbetween Yfh1p and ferrochelatase was measured when Yfh1pwas used as the analyte (Fig. 9). Varying the concentration ofYfhp1 from 7� 10�8 to 1.4� 10�6

M allowed us to estimatethe dissociation constant (4� 10�8

M) by global fitting to asimple Langmuir model.

Figure 6. Suppression of the cytochrome-deficient phenotype of Dyfh1 aftergrowth on a non-fermentable carbon source. Freshly isolated colonies ofW303Dyfh1 (on YPG-agar) were grown in either YPR or YPG liquid medium.Low-temperature spectra of whole cells were recorded after 24 h on YPR (A) or5 days on YPG (B). Cells from this last culture were plated on YPD-agar, andthen re-inoculated on YPR liquid medium. The low-temperature spectrum ofwhole cells was recorded after 24 h (C).

Figure 7. Apparition of suppressor colonies on YPG-agar plate. A cell lawn ofW303Dyfh1 on YPD-agar was replicated on YPG-agar and incubated for 5 daysat 30�C. Suppressor colonies grew as big red colonies (phenotype of respira-tory-competent ade2 cells) on a white background of slow-growing cells(respiratory-deficient).




Dow

nloaded from


DISCUSSION

The loss-of-function of YFH1 results in striking maldistributionof cellular iron, with depletion of cytoplasmic iron andaccumulation of mitochondrial iron. Here we show thatmitochondrial iron in Dyfh1 cells, although present in largeexcess, is unavailable for haeme synthesis. Ferrochelatase, amitochondrial protein, able to mediate insertion of iron or zincinto the porphyrin precursor, makes primarily the zincprotoporphyrin product. Zinc protoporphyrin instead of haemeaccumulates during growth of Dyfh1 mutant cells and,furthermore, preferential formation of zinc protoporphyrin isobserved in real time. The method for these studies involvesdirect presentation of porphyrin to mitochondria and toferrochelatase of permeabilized cells with intact architecture,thereby specifically testing the iron delivery portion of thehaeme biosynthetic pathway. The studies show that Dyfh1mutant cells exhibit a defect in iron use by ferrochelatase.

Multiple secondary effects and the complex pleiotropicphenotype of these cells has impeded definition of the functionof frataxins. The defect in haeme synthesis in Yfh1p depletedcells, however, is readily reversed by reinduction of expressionfrom a regulated promoter, and therefore the phenotype is notdue to secondary effects such as mtDNA damage. The precise

role of Yfh1p and frataxins in iron delivery for haeme synthesisremains to be defined. A role for Yfh1 in iron solubilization inmitochondria is possible and/or a role in directly handing of ironas described for metallation of copper proteins by copperchaperones (29). Ferrochelatase is able to use soluble ferrousiron as a substrate for iron incorporation into PPIX, and iron islargely present as insoluble ferric particles in Dyfh1 mitochon-dria. This could explain why the large excess of iron in thesemitochondria is not available for haeme synthesis, and maybe forother iron-requiring processes like Fe–S centre assembly. Theinvolvement of Yfh1p might involve a role in maintainingsolubility of iron in mitochondria. Our results show thatHem15p and Yfh1p physically interact in vitro. An in vivointeraction between these two proteins might mediate hand-offof iron for haeme formation. Adamec et al. (30) showed thatself-assembled multimers of Yfh1p can sequester more than3000 atoms of iron, and that iron can be released from theprotein shell by a reducing agent. It is tempting to speculate thatiron bound to such a Yfh1p intermediate may be the iron donorto ferrochelatase for haeme synthesis. Such a donor should existas it is unlikely that there is ‘free’ iron in mitochondria, which isan environment particularly sensitive to radical reactions.

The role of Yfh1p in iron delivery for haeme synthesis recallsthe recently described role of Yfh1p in iron delivery for Fe–S

Figure 8. Phenotype of a suppressor strain isolated from a Dyfh1 strain (X498-1A). 1, WT (YPH499); 2, Dyfh1 (X498-1A); 3, suppressor strain isolated fromX498-1A (sup4þ). (A) Low temperature spectra of whole cells. (B) Aconitase activity of isolated mitochondrial fractions. (C) Blotting of Hem15p and Cyc1p(1, WT; 2, Dyfh1; 3, sup4þ; 4, diploid from the cross sup4þ X 5Cyfh1; 5–8, tetrad from the diploid). (D) Oxygen uptake by whole cells grown overnight onYPR and incubated in 0.1 M K-phosphate buffer (pH 7) containing 2% glucose. (E) Iron accumulation by whole cells grown for 16 h on YPD þ various concen-trations of Fe(III)-citrate. (F) Growth of cells on YPD-agar plates containing 500mM ferric citrate (A), 1 mM H2O2 (B) or 2 mM CuSO4 (C). When present, errorbars represent SE from three experiments.




Dow

nloaded from


cluster synthesis in mitochondria (8,15–18). The precisefunction of Yfh1p is likewise undefined in this process. Arecent study by Puccio et al. (17) analysed the sequence ofevents in tissues of transgenic mice following frataxindepletion. They found that the Fe–S assembly defect appearedprior to mitochondrial iron accumulation, suggesting a primaryrole for frataxin in iron utilization. Our work also indicates thatYfh1 is involved in mitochondrial iron use (for haeme synthesisas for Fe–S cluster synthesis) and for correct partitioning ofiron between mitochondrial and cytoplasmic pools. Definitionof the mechanisms of these effects will require additional work,and the identification of additional components of themitochondrial iron transport system(s).

The cytochrome-deficient phenotype of Dyfh1 cells is oftenmasked by nuclear mutation(s) that suppress the phenotype.The features of strains carrying such suppressor mutations arestriking. Haeme synthesis is restored as indicated by completerecovery of cytochrome spectra and disappearance of Zn-PPIX.Haeme synthesis is restored, although iron accumulation andFe–S cluster deficiency phenotypes are intermediate betweenthe wild-type and deletion phenotypes. We frequently encoun-tered such genetic changes with various Dyfh1 strains: strainswere cytochrome-deficient when freshly isolated and acquirednormal pigmentation after a few culture cycles on completemedium. Our finding that Dyfh1 cells accumulate suppressormutation(s) fits well with a recent work (19), showing that theabsence of Yfh1 in yeast leads to nuclear damage, increasedchromosomal instability including recombination and muta-tion, and greater sensitivity to DNA-damaging agents. Thefrequency of these events may explain why the link of Yfh1pwith cytochromes and haeme synthesis has not previously beendescribed. The implication of the results is that the requirementof Yfh1p in haeme synthesis can be bypassed by nuclear genemutations. Identifying the suppressor mutation(s) that allow(s)Dyfh1 cells to recover normal cytochrome synthesis andrespiration is being undertaken. The unidentified suppressormutation(s) might result in solubilizing part of the ironaccumulated into Dyfh1 mitochondria, making it available forhaeme synthesis, as suggested by our Mossbauer analysis. Thelow expression of frataxin in erythroid cells (11) and the lack ofred cell phenotypes in the human frataxin deficiency disease

Friedreich ataxia (11) likewise suggest that haeme synthesiscan occur in the absence of frataxin in some settings.

Ferrochelatases from diverse species as mammals,Drosophila, Schizosaccharomyces pombe and some bacteriapossess 2Fe–2S clusters that are critical for activity (reviewedin 31). In view of the role of frataxin in Fe–S cluster formation,an effect of frataxin deficiency on ferrochelatase function inthese species might be inferred. By contrast, the ferrochelataseof S. cerevisiae lacks such as Fe–S cluster and appears to be anoutlier in this regard. One can speculate about the implicationsof this difference for the regulatory links between frataxin andferrochelatase function. Perhaps, the effects of frataxin on Fe–Scluster assembly being insufficient to inactivate ferrochelatasein S. cerevisiae, alternative regulatory mechanisms evolvedinstead. At least two such mechanisms are observed here. First,a transcriptional mechanism is implied by the decreasedHem15 transcript levels observed in the absence of Yfh1.The mediators of this response are not Aft1 or Aft2, thepreviously identified iron regulatory proteins (22,23). Second apost-transcriptional effect of Yfh1 is required for activeferrochelatase function. This might involve a role of Yfh1 indelivery of the substrate iron for haeme formation. Alternativepossibilities are that Yfh1 activates ferrochelatase activity byproducing a folding or conformational change in the protein, orby contribution of a catalytic iron atom that serves a functionanalogous to the regulatory Fe–S cluster. The direction ofthese regulatory effects makes sense in terms of couplinghaeme synthesis to the availability of the critical iron cofactor.In similar fashion, the first step in haeme biosynthesis,d-aminolevulinate synthase in erythroid cells, is positivelyregulated by iron (32) and ferrochelatase in mammals is alsoregulated by iron via post-transcriptional mechanisms (33).Many questions remain: the molecular details of the regulatoryeffects of Yfh1 on ferrochelatase function remain to be defined.Also a question exists as to whether similar control mechan-isms will be found in mammalian tissues.

MATERIALS AND METHODS

Strains, media and growth conditions

The strains of S. cerevisiae used in this study are described inTable 1. Unless otherwise stated, cells were grown at 30�C incomplete YPD medium supplemented with 15 mg hemin/ml(YPD-hemin) for haeme-deficient strains. Other media usedwere YPG (yeast extract 1%, peptone 1%, glycerol 2%), YPR(yeast extract 1%, peptone 1%, raffinose 2%) and YPGal (yeastextract 1%, peptone 1%, galactose 2%). Cultures in iron-deficient medium were done in minimal YNB-glucose medium(yeast nitrogen base without copper and iron, Bio 101, Inc.)plus the required amino acids and 1 mM copper sulfate.

Iron uptake assays

Iron uptake was measured in microtitre plates. Cells weregrown overnight at 30�C in YPD-hemin to the stationary phase.Cells were diluted 10-fold in the same medium and cultured for6 h at 30�C. They were then washed once with 2% EDTA andthree times with distilled water and suspended in 50 mM citrate

Figure 9. Yfh1 and Hem15 interact in vitro. Typical sensorgram obtained whenmeasuring the interaction of 7� 10�8

M purified Yfh1p in HBS without EDTAinjected (1) over 100 RU immobilized ferrochelatase (A) or 100 RU immobi-lized BSA (B). Dissociation of the Yfh1p–Hem15p complex was analysedafter removal of Yfh1p from the mobile phase (2). The solid line over curve(A) represents the best fit of experimental data with a 1:1 Langmuir modelof interaction.




Dow

nloaded from


(tri-sodium) buffer (pH 6.5) containing 5% glucose to giveabout 2 mg protein/ml. The cell suspension was distributedinto the wells of the micro-titration plate (50 ml cells/well)at 0�C. Iron was added [as 55Fe(II)-ascorbate] to give afinal concentration of 1–5 mM and the plate was incubated for15–60 min at 30�C. The cells were collected with a cellharvester (Brandel) and washed on the filter.

RNA isolation and northern analysis

RNA was extracted as described previously (34). Northernblotting and hybridization (at 42�C in 50% [vol/vol] forma-mide) were done essentially as described previously (35). TheDNA fragment used as a probe for HEM15 was a 1.2 kbEcoRI–BamHI fragment and, for ACT1, a 1.2 kb BamHI–HindIII fragment.

Biacore experiments

Recombinant yeast ferrochelatase was overproduced inEscherichia coli (36) and purified as previously described (21).The specific activity of the recombinant protein (35 000 nmol/h/mg protein) was similar to that of the native enzyme.Recombinant mature Yfh1p (amino acids 52–174) was over-produced in E. coli as a His-tagged protein (37). Both proteinswere extensively dialysed against HBS-buffer [10 mM HEPES,150 mM NaCl and 0.005% (v/v) Tween 20, pH 7.4].

Mossbauer spectroscopy

In situ Mossbauer spectroscopy enables in principle thesimultaneous identification of all major iron metabolites on aqualitative as well as on a quantitative level and of the kineticsof intracellular iron distribution without destruction of thecellular assembly (38,39). A prerequisite for in situ spectro-scopy is an appropriate feeding of the cell cultures with 57Feleading to 57Fe-iron concentrations high enough for detectionby Mossbauer spectroscopy. 57Fe was purchased from

Physikalische Messtechnik Lubeck. For preparation ofMossbauer samples cells were isotopically enriched with 57Feby adding an aqueous 57Fe(III)–citrate solution (10 mM final) tothe medium. As later demonstrated by Mossbauer spectro-scopy, the intracellular isotopic enrichment with 57Fe wassufficiently high to obtain well-resolved spectra.

Samples of isolated mitochondria were transferred tocylindrical Mossbauer sample holders made of Delrin1, frozenand stored at 77 K until measurement. The Mossbauer spectro-meter worked in conventional constant acceleration mode withsources of 0.9–1.85 GBq 57Co/Rh (Techsnabexport). Thespectrometer was calibrated against a metallic a-iron foil atroom temperature yielding a standard line width of 0.25 mm/s.The Mossbauer cryostats were a helium bath cryostat (MD306,Oxford Instruments) or a superconducting magnet system withsplit coil geometry (Oxford Instruments). Isomer shift, quadru-pole splitting, EQ, and percentage of the total absorption areawere obtained by least squares fits of Lorentzian lines to theexperimental spectra.

Analytical methods

Protein concentrations were determined based on the Lowrymethod with the Bio-Rad Dc protein assay and the amount ofiron was determined with the nitro-PAPS assay (40).Polyacrylamide gel electrophoresis (PAGE) was carried outby the Laemmli procedure (41). For SDS–PAGE the ready-to-use NuPAGE1 electrophoresis system (Invitrogen) was used.Proteins were visualized by Coomassie brilliant blue staining.Iron loading of the protein could be detected on gels using thespecific iron staining method of Fridovich (42). Phosphate wasdetermined employing literature methods (43,44).

Other

Protoplasts were lysed and fractionated to purify isolatedmitochondria as previously described (45). Low temperaturespectra (�191�C) of whole cells were prepared as described

Table 1. Genotype of the Saccharomyces cerevisiae strains used

Strain name Genotype

S150-2B MATa, his3-D1, leu2-3,112, trp1-289, ura3-52S150-2Br0 MATa, his3-D1, leu2-3,112, trp1-289, ura3-52, rho�

S150-2BDhem1 MATa, his3-D1, leu2-3,112, trp1-289, ura3-52, hem1D::LEU2S150-2BDhem13 MATa, his3-D1, leu2-3,112, trp1-289, ura3-52, hem13D1::URA3S150-2BDhem14 MATa, his3-D1, leu2-3,112, trp1-289, ura3-52, hem14D:: kanMX4S150-2BDhem15 MATa, his3-D1, leu2-3,112, trp1-289, ura3-52, hem15D::hisGS150-2BDyfh1 MATa, his3-D1, leu2-3,112, trp1-289, ura3-52, yfh1D::TRP1Y18 MATa, trp1-63, leu2-3, 112, gcn4-101, his3-609, aft1::TRP1Y18Daft2 MATa, trp1-63, leu2-3, 112, gcn4-101, his3-609, aft1D::TRP1, aft2D::kanMX4CM3260 MATa, trp1-63, leu2-3, 112, gcn4-101, his3-609CM3260Daft2 MATa, trp1-63, leu2-3, 112, gcn4-101, his3-609, aft2::kanMX4W303-1A MATa, ade2-1, his3-11,15, leu2-3,112, trp1-1, ura3-1, can1-100W303Dyfh1 MATa, ade2-1, his3-11,15, leu2-3,112, trp1-1, ura3-1, can1-100, yfh1D::TRP1W303Dyfh1sup3 MATa, ade2-1, his3-11,15, leu2-3,112, trp1-1, ura3-1, can1-100, yfh1D::TRP1, Yfh1sup3YPH499 MATa, ura3-52, lys2-801, ade2-101, trp1D63, his3D200, leu2D1X498-1A MATa, ura3-52, lys2-801, ade2-101, trp1D63, his3D200, leu2D1, yfh1D::TRP1Sup4þ MATa, ura3-52, lys2-801, ade2-101, trp1D63, his3D200, leu2D1, yfh1D::TRP1, Yfh1sup4GAL-Yfh1 MATa, ura3-52, lys2-801, ade2-101, his3D200, cyh2, yfh1D::TRP1, Gal-Yfh1::URA35Cyfh1 MATa, yfh1D::HIS3, ade2-1, leu2-3,112, trp1, ura3-52




Dow

nloaded from


previously (46). Total haeme content of the cells was determinedby the pyridine hemochromogen method (47). Zinc contentof the cells was measured by inductively coupled plasmaatomic emission spectroscopy. Antibodies against Yfh1p,Hem15p and Hem13p were obtained as previously described(21,48,49). Ferrochelatase, zinc-chelatase and protopor-phyrinogen oxidase activities were measured fluorimetrically(21,50). Protoporphyrinogen was prepared from protoporphyrinby reduction with a sodium amalgam (48). When required, cellswere made permeable by adding 20 ml of toluene–ethanol (4:1)to 1 ml of cell suspension (OD 10). Aconitase activity wasmeasured spectrophotometrically on mitochondrial extracts ofcells grown in YP with 2% raffinose (51).

ACKNOWLEDGEMENTS

R.S. is supported by the Association Francaise de l’Ataxie deFriedreich. This work was supported by grants from the FrenchMinistere de la Recherche (Programme de RecherchesFondamentales en Microbiologie, Maladies Infectieuses etParasitaires and Reseau Infection Fongique) and from theAssociation pour la Recherche sur le Cancer (ARC 5439). Thiswork was supported by National Institutes of Health GrantDK53953 (to A.D.). We thank Debkumar Pain for manyhelpful discussions, strains and plasmids.

REFERENCES

1. Campuzano, V., Montermini, L., Molto, M.D., Pianese, L., Cossee, M.,Cavalcanti, F., Monros, E., Rodius, F., Duclos, F., Monticelli, A. et al.(1996) Friedreich’s ataxia: autosomal recessive disease caused by anintronic GAA triplet repeat expansion. Science, 271, 1423–1427.

2. Puccio, H. and Koenig, M. (2000) Recent advances in the molecularpathogenesis of Friedreich ataxia. Hum. Mol. Genet., 9, 887–892.

3. Babcock, M., de Silva, D., Oaks, R., Davis-Kaplan, S., Jiralerspong, S.,Montermini, L., Pandolfo, M. and Kaplan, J. (1997) Regulation ofmitochondrial iron accumulation by Yfh1p, a putative homolog of frataxin.Science, 276, 1709–1712.

4. Foury, F. and Cazzalini, O. (1997) Deletion of the yeast homologue of thehuman gene associated with Friedreich’s ataxia elicits iron accumulation inmitochondria. FEBS Lett., 411, 373–377.

5. Koutnikova, H., Campuzano, V., Foury, F., Dolle, P., Cazzalini, O. andKoenig, M. (1997) Studies of human, mouse and yeast homologues indicatea mitochondrial function for frataxin. Nat. Genet., 16, 345–351.

6. Lodi, R., Cooper, J.M., Bradley, J.L., Manners, D., Styles, P., Taylor, D.J.and Schapira, A.H. (1999) Deficit of in vivo mitochondrial ATP productionin patients with Friedreich ataxia. Proc. Natl Acad. Sci. USA, 96,11492–11495.

7. Ristow, M., Pfister, M.F., Yee, A.J., Schubert, M., Michael, L., Zhang, C.Y.,Ueki, K., Michael, M.D., Lowell, B.B. and Kahn, C.R. (2000) Frataxinactivates mitochondrial energy conversion and oxidative phosphorylation.Proc. Natl Acad. Sci. USA, 97, 12239–12243.

8. Rotig, A., de Lonlay, P., Chretien, D., Foury, F., Koenig, M., Sidi, D.,Munnich, A. and Rustin, P. (1997) Aconitase and mitochondrial iron-sulphur protein deficiency in Friedreich ataxia. Nat. Genet., 17, 215–217.

9. Bradley, J.L., Blake, J.C., Chamberlain, S.,Thomas, P.K., Cooper, J.M.and Schapira, A.H. (2000) Clinical, biochemical and molecular geneticcorrelations in Friedreich’s ataxia. Hum. Mol. Genet., 9, 275–282.

10. Wong, A., Yang, J., Cavadini, P., Gellera, C., Lonnerdal, B., Taroni, F. andCortopassi, G. (1999) The Friedreich’s ataxia mutation confers cellularsensitivity to oxidant stress which is rescued by chelators of iron andcalcium and inhibitors of apoptosis. Hum. Mol. Genet., 8, 425–430.

11. Becker, E.M., Greer, J.M., Ponka, P. and Richardson, D.R. (2002) Erythroiddifferentiation and protoporphyrin IX down-regulate frataxin expression inFriend cells: characterization of frataxin expression compared to moleculesinvolved in iron metabolism and hemoglobinization. Blood, 99, 3813–3822.

12. Radisky, D.C., Babcock, M.C. and Kaplan, J. (1999) The yeast frataxinhomologue mediates mitochondrial iron efflux. Evidence for a mitochon-drial iron cycle. J. Biol. Chem., 274, 4497–4499.

13. Wilson, R.B. and Roof, D.M. (1997) Respiratory deficiency due to loss ofmitochondrial DNA in yeast lacking the frataxin homologue. Nat. Genet.,16, 352–357.

14. Foury, F. and Talibi, D. (2001) Mitochondrial control of iron homeostasis.A genome wide analysis of gene expression in a yeast frataxin-deficientstrain. J. Biol. Chem., 276, 7762–7768.

15. Duby, G., Foury, F., Ramazzotti, A., Herrmann, J. and Lutz, T. (2002)A non-essential function for yeast frataxin in iron–sulfur cluster assembly.Hum. Mol. Genet., 11, 2635–2643.

16. Muhlenhoff, U., Richhardt, N., Ristow, M., Kispal, G. and Lill, R. (2002)The yeast frataxin homolog Yfh1p plays a specific role in the maturation ofcellular Fe/S proteins. Hum. Mol. Genet., 11, 2025–2036.

17. Puccio, H., Simon, D., Cossee, M., Criqui-Filipe, P., Tiziano, F., Melki, J.,Hindelang, C., Matyas, R., Rustin, P. and Koenig, M. (2001) Mouse modelsfor Friedreich ataxia exhibit cardiomyopathy, sensory nerve defect andFe–S enzyme deficiency followed by intramitochondrial iron deposits. Nat.Genet., 27, 181–186.

18. Puccio, H. and Koenig, M. (2002) Friedreich ataxia: a paradigm formitochondrial diseases. Curr. Opin. Genet. Dev., 12, 272–277.

19. Karthikeyan, G., Lewis, L.K. and Resnick, M.A. (2002) The mitochondrialprotein frataxin prevents nuclear damage. Hum. Mol. Genet., 11,1351–1362.

20. Branda, S.S., Yang, Z.Y., Chew, A. and Isaya, G. (1999) Mitochondrialintermediate peptidase and the yeast frataxin homolog together maintainmitochondrial iron homeostasis in Saccharomyces cerevisiae. Hum. Mol.Genet., 8, 1099–1110.

21. Camadro, J.M. and Labbe, P. (1988) Purification and properties offerrochelatase from the yeast Saccharomyces cerevisiae. Evidence for aprecursor form of the protein. J. Biol. Chem., 263, 11675–11682.

22. Blaiseau, P.-L., Lesuisse, E. and Camadro, J.M. (2001) Aft2p, a noveliron-regulated transcription activator that modulates, with Aft1p,intracellular iron use and resistance to oxidative stress in yeast. J. Biol.Chem., 276, 34221–34226.

23. Yamaguchi-Iwai, Y., Dancis, A. and Klausner, R.D. (1995) AFT1: amediator of iron regulated transcriptional control in Saccharomycescerevisiae. EMBO J., 14, 1231–1239.

24. Bauminger, E.R., Cohen, S.G., Dickson, D.P., Levy, A., Ofer, S. andYariv, J. (1980) Mossbauer spectroscopy of Escherichia coli and itsiron-storage protein. Biochim. Biophys. Acta, 623, 237–242.

25. Bauminger, E.R., Cohen, S.G., Labenski de Kanter, F., Levy, A., Ofer, S.,Kessel, M. and Rottem, S. (1980) Iron storage in Mycoplasma capricolum.J. Bacteriol., 141, 378–381.

26. Bauminger, E.R., Harrison, P.M., Nowik, I. and Treffry, A. (1989)Mossbauer spectroscopic study of the initial stages of iron-core formationin horse spleen apoferritin: evidence for both isolated Fe(III) atoms andoxo-bridged Fe(III) dimers as early intermediates. Biochemistry, 28,5486–5493.

27. St Pierre, T.G., Bell, S.H., Dickson, D.P., Mann, S., Webb, J., Moore, G.R.and Williams, R.J. (1986) Mossbauer spectroscopic studies of the cores ofhuman, limpet and bacterial ferritins. Biochim. Biophys. Acta, 870,127–134.

28. Foury, F. and Roganti, T. (2002) Deletion of the mitochondrial carrier genesMRS3 and MRS4 suppresses mitochondrial iron accumulation in a yeastfrataxin-deficient strain. J. Biol. Chem., 277, 24475–24483.

29. Portnoy, M.E., Schmidt, P.J., Rogers, R.S. and Culotta, V.C. (2001) Metaltransporters that contribute copper to metallochaperones in Saccharomycescerevisiae. Mol. Genet. Genom., 265, 873–882.

30. Adamec, J., Rusnak, F., Owen, W.G., Naylor, S., Benson, L.M., Gacy, A.M.and Isaya, G. (2000) Iron-dependent self-assembly of recombinant yeastfrataxin: implications for Friedreich ataxia. Am. J. Hum. Genet., 67,549–562.

31. Ferreira, G.C. (1999) Ferrochelatase. Int. J. Biochem. Cell Biol., 31,995–1000.

32. Ponka, P. (1997) Tissue-specific regulation of iron metabolism and hemesynthesis: distinct control mechanisms in erythroid cells. Blood, 89, 1–25.

33. Taketani, S., Adachi, Y. and Nakahashi, Y. (2000) Regulation of theexpression of human ferrochelatase by intracellular iron levels. Eur. J.Biochem., 267, 4685–4692.

34. Kohrer, K. and Domdey, H. (1991) Preparation of high molecular weightRNA. Meth. Enzymol., 194, 398–405.




Dow

nloaded from


35. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: aLaboratory Manual, 2nd edn. Cold Spring Harbor Laboratory, Cold SpringHouston, NY.

36. Gora, M., Chacinska, A., Rytka, J. and Labbe-Bois, R. (1996) Isolation andfunctional characterization of mutant ferrochelatases in Saccharomycescerevisiae. Biochimie, 78, 144–152.

37. Gordon, D.M., Wang, J., Amutha, B. and Pain, D. (2001) Self-associationand precursor protein binding of Saccharomyces cerevisiae Tom40p, thecore component of the protein translocation channel of the mitochondrialouter membrane. Biochem. J., 356, 207–215.

38. Matzanke, B.F. (1987) Mossbauer spectroscopy of microbial iron uptakeand metabolism. In Winkelmann, G., van der Helm, D. and Neilands, J.B.(eds), Iron Transport in Microbes, Plants and Animals. Verlag Chemie,Weinheim, pp. 251–284.

39. Matzanke, B.F. (1991) Structure, coordination chemistry and functions ofmicrobial iron chelates. In Winkelmann, G. (ed.), Handbook ofMicrobial Iron Chelates (Siderophores). CRC Press, Boca Raton, FL,pp. 15–60.

40. Makino, T., Kiyonaga, M. and Kina, K. (1988) A sensitive, directcolorimetric assay of serum iron using the chromogen, nitro-PAPS. Clin.Chim. Acta, 171, 19–27.

41. Laemmli, U.K. (1970) Cleavage of structural proteins during the assemblyof the head of bacteriophage T4. Nature, 227, 680–685.

42. Kuo, C.F. and Fridovich, I. (1988) A stain for iron-containing proteinssensitive to nanogram levels of iron. Anal. Biochem., 170, 183–185.

43. Chen, P.S., Toribara, T.Y. and Warner, H. (1956) Microdetermination ofphosphorus. Anal. Chem., 28, 1756–1758.

44. Fiske, C.H. and Subbarow, Y.P. (1925) The colorimetric determination ofphosphorus. J. Biol. Chem., 66, 375–400.

45. Raguzzi, F., Lesuisse, E. and Crichton, R.R. (1988) Iron storage inSaccharomyces cerevisiae. FEBS Lett., 231, 253–258.

46. Lesuisse, E. and Labbe, P. (1989) Reductive and non-reductive mechanismsof iron assimilation by the yeast Saccharomyces cerevisiae. J. Gen.Microbiol., 135, 257–263.

47. Porra, R.J. and Jones, O.T. (1963). Studies on ferrochelatase. 1. Assay andproperties of ferrochelatase from a pig-liver mitochondrial extract.Biochem. J., 87, 181–185.

48. Camadro, J.M., Chambon, H., Jolles, J. and Labbe, P. (1986) Purificationand properties of coproporphyrinogen oxidase from the yeastSaccharomyces cerevisiae. Eur. J. Biochem., 156, 579–587.

49. Gordon, D.M., Kogan, M., Knight, S.A., Dancis, A. and Pain, D. (2001)Distinct roles for two N-terminal cleaved domains in mitochondrial importof the yeast frataxin homolog, Yfh1p. Hum. Mol. Genet., 10, 259–269.

50. Camadro, J.M., Thome, F., Brouillet, N. and Labbe, P. (1994) Purificationand properties of protoporphyrinogen oxidase from the yeastSaccharomyces cerevisiae. Mitochondrial location and evidence for aprecursor form of the protein. J. Biol. Chem., 269, 32085–32091.

51. Schilke, B., Voisine, C., Beinert, H. and Craig, E. (1999) Evidence for aconserved system for iron metabolism in the mitochondria ofSaccharomyces cerevisiae. Proc. Natl Acad. Sci. USA, 96, 10206–10211.




Dow

nloaded from
