Human Molecular Genetics

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (30) Request Permissions Google Scholar Articles by Tan, G. Articles by Cortopassi, G. Search for Related Content PubMed PubMed Citation Articles by Tan, G. Articles by Cortopassi, G. Social Bookmarking          What's this?

Human Molecular Genetics, 2003, Vol. 12, No. 14 1699-1711
DOI: 10.1093/hmg/ddg187
© 2003 Oxford University Press

Decreased expression of genes involved in sulfur amino acid metabolism in frataxin-deficient cells

Guolin Tan¹, Eleonora Napoli¹, Franco Taroni² and Gino Cortopassi^1,*

¹Department of Molecular Biosciences, 1311 Haring Hall, University of California, Davis, CA 95616, USA and ²Division of Biochemistry and Genetics, Istituto Nazionale Neurologico ‘Carlo Besta’, Via Celoria 11, Milan, Italy

Received March 11, 2003; Accepted May 16, 2003

	ABSTRACT

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIAL AND METHODS REFERENCES

 
Inherited deficiency of the mitochondrial protein frataxin causes neural and cardiac cell degeneration, and Friedreich's ataxia. Five hypotheses for frataxin's mitochondrial function have been generated, largely from work in non-human cells: iron transporter, iron–sulfur cluster assembler, iron-storage protein, antioxidant and stimulator of oxidative phosphorylation. We analyzed gene expression in three human cell types using microarrays, and identified just 48 transcripts whose expression was significantly frataxin-dependent in at least two cell types. Significant decreases in seven transcripts occurred in the sulfur amino acid (SAA) biosynthetic pathway and the iron–sulfur cluster (ISC) biosynthetic pathway to which it is connected. By contrast, we did not observe a single frataxin-dependent transcript that fits with the other four current hypotheses. Quantitative reverse-transcriptase PCR analysis of ISC-S and rhodanese transcripts confirmed that the expression of these genes involved in ISC metabolism was lower in mutants. Amino acid analysis confirmed the defect in SAA metabolism: homocystine, cysteine, cystathionine and serine were significantly decreased in frataxin-deficient cell extracts and mitochondria. An ISC defect was further confirmed by observing decreases in succinate dehydrogenase and aconitase activities, whose activities require ISCs. The ISC-U scaffold protein was specifically decreased in frataxin-deficient cells, suggesting a role for frataxin in its expression or maintenance, and sodium sulfide partially rescued the oxidant-sensitivity of the FRDA cells. Also, multiple transcripts involved in the Fas/TNF/INF apoptosis pathway were up-regulated in frataxin-deficient cells, consistent with a multi-step mechanism of Friedreich's ataxia pathophysiology, and suggesting alternative possibilities for therapeutic intervention.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIAL AND METHODS REFERENCES

 
Friedreich's ataxia (FRDA) is an autosomal-recessive neuro- and cardio-degenerative disease characterized by progressive ataxia and cardiomyopathy. FRDA is the most common inherited ataxia, with a prevalence of 1/50 000 individuals (1), and is usually caused by inheritance of two intronic expansions of (GAA)_>120 in the frataxin gene, which inhibit expression of the frataxin protein (2). In spite of a large number of studies which have focused on the mitochondrial function of frataxin protein in bacteria, yeast, mice and humans, this function is still controversial (3–6).

At least five hypotheses for the primary mitochondrial function of frataxin have been proposed: iron transport (3), ISC biosynthesis (4,7), iron storage (8), antioxidant (5) and stimulator of oxidative phosphorylation (6).

In the yeast S. cerevisiae, deletion of frataxin homolog 1 (YFH1) produces a 10-fold increase in mitochondrial iron and increased sensitivity to oxidants (3,9). Further, in human FRDA patients excess iron is observed in hearts and neural tissue (10,11). These data suggested a role of frataxin in mitochondrial iron homeostasis, and from them a role of frataxin as a mitochondrial iron exporter was proposed. Subsequent work with mice completely deficient in frataxin demonstrated cardiodegeneration and iron accumulation (12). However, subsequent studies of mitochondrial iron concentration in human cells suggested only a modest (13) or not significant increase in total mitochondrial iron (14,15), inconsistent with the hypothesis that frataxin is a mitochondrial iron exporter.

A second hypothesis for frataxin function is in the biogenesis of iron–sulfur clusters (ISCs). It was shown that yeast bearing the YFH1 mutation were deficient in multiple ISC-dependent enzyme activities (4). Furthermore a selective deficiency of enzyme activities that require ISC was demonstrated by cardiac biopsy of an FRDA patient (7). Finally, although total deletion of the Frda gene in the mouse resulted in embryonic lethality, with evidence of iron accumulation (12), subsequent studies with conditional knockout mice demonstrated that the defects in mitochondrial ISC enzyme activity precede iron accumulation (16), suggesting that iron accumulation is a distal consequence of an earlier, proximal consequence of frataxin deficiency. Studies have now clearly established that mitochondrial iron accumulation is a general consequence of deficiencies in ISC biogenesis in yeast models (17–19). Recently, deficiency of the yeast frataxin homolog protein yfh1p has been demonstrated to cause a partial defect in the maturation of mitochondrial iron–sulfur clusters in yeast (20,21).

Three other hypotheses for the function of frataxin have received experimental support, including that frataxin is: a mitochondrial iron-storage protein (22), a mitochondrial antioxidant (5,23), and a general stimulator of oxidative phosphorylation (6).

As most of the five hypotheses have been substantially motivated by results in non-human systems, and FRDA is a human disease, it is important to identify which of frataxin's potential functions is the most likely in human cells, and how this primary deficiency triggers the pathophysiological steps which ultimately cause cell degeneration, because each of these consequent steps may themselves be avenues for therapeutic intervention (15). Thus, using fibroblast and lymphoblast cells from the FRDA patients themselves, and a neural NT2 cell line in which frataxin expression has been knocked down by RNAi inhibition as a cell model of a ‘target’ neural tissue, we have carried out a microarray analysis of the specific effects of frataxin deficiency in human cells. These results have only supported one of the five current hypotheses, i.e. a role of frataxin in ISC biogenesis or maintenance, to the extent that we observe multiple transcriptional alterations in the ISC sulfur amino acid (SAA) pathway. Five independent molecular, biochemical and cellular tests each confirm a frataxin-dependent alteration of ISC/SAA biogenesis, including QRT–PCR of ISC-related genes, amino acid analysis, western analysis of ISC-U expression, ISC-dependent enzyme analysis, and sulfide-dependent rescue of the major cellular FRDA phenotype.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIAL AND METHODS REFERENCES

 
Development of a neuronal model of FRDA by frataxin^RNAi transfection
Available FRDA cellular models for FRDA consist of patient-derived fibroblast or lymphoblastoid cells. In order to generate a cellular model of a neural lineage, e.g. one which degenerates in FRDA, human neuronal precursor NT2 cells were transfected with frataxin-specific interfering RNA. RNA interference using a short interfering RNA (called RNAi or siRNA) is a recently developed technique with the potential for strong inhibition of a targeted message (24). Thus human neuronal precursor NT2 cells were transfected with Frataxin^RNAi. This resulted in a >70% reduction of the frataxin mRNA, by quantitative RT–PCR, relative to transfection with the ‘scrambled’ RNAi, i.e. an RNAi of identical nucleotide content but in randomized order (Fig. 1A). The significance and extent of the reduction in frataxin gene expression was further confirmed at the protein level by western blot (Fig. 1B).

View larger version (26K): [in this window] [in a new window]   Figure 1. Frataxin expression levels in neural cells (NT2) transfected with scrambledRNAi or frataxinRNAi. (A) NT2 cells were transfected with frataxinRNAior scrambledRNAi, mRNA prepared and analyzed by quantitative RT–PCR, using primers specific to the frataxin gene, error bars represent standard deviation of three experiments. (B) Western blot detection of frataxin protein, using anti-frataxin antibody, 48 h post-transfection with scrambledRNAi (lanes 1, 3 and 5) or frataxinRNAi (lanes 2, 4 and 6).

 
Mutant fibroblasts, lymphoblasts and frataxin-inhibited NT2s have consistent alterations in gene expression
RNA was prepared from six fibroblast lines (three mutants; three controls), eight lymphoblast lines (three mutants; three controls; two mutants transfected with frataxin), and six neural cell lines (three NT2 controls, three NT2 transfected with frataxin^RNAi), labeled and hybridized to the Affymetrix U95Av2 human chip, on which are represented 12 599 genes. Chips were evaluated for overall labeling quantity and quality, and any chips with significant aberrations were excluded. Raw Affymetrix chip data (i.e. cel files) were analyzed using the DNA-chip data analyzer (25), which included normalization by total chip fluorescence. The first criterion for comparison was the absent or present call, determined by the Affymetrix software. Absent calls are those transcripts whose expression is so low that they are essentially below the level of detection. As can be seen in Table 1, in each cell type about half of the transcripts were detected as ‘present’ at an intensity sufficient for further analysis, which is the usual result in our experience, although the transcript identities are somewhat different between cell types, as is expected because different cell types express transcripts to different extents. A second convenient criterion for gene expression changes used in the microarray field is the 1.5-fold change level, i.e. a relative expression from control level of 50% increase or 50% decrease or larger in amplitude. Using these criteria, 645 changes were detected in mutant lymphoblasts, 468 in mutant fibroblasts, and 185 in NT2 cells transfected with frataxin^RNAi. Consistently, the largest variation between mutants and controls occurred in the order lymphoblasts>fibroblasts>NT2 cells transfected with frataxin^RNAi, which is probably a consequence of the EBV transfection of the lymphoblasts that could result in aneuploidy, and the nuclear genetic heterogeneity of both the lymphoblasts and fibroblasts, relative to the uniform nuclear background of the NT2 cells.

View this table: [in this window] [in a new window]   Table 1. Microarray parameters

 
An attempt to ‘trap’ frataxin-dependent transcripts using cell autonomous gene expression
Our goal was to identify the frataxin-dependent transcriptional variation by microarray analysis, i.e. a frataxin-dependent ‘signal’ in a background of transcriptional ‘noise’. The noise, i.e. the frataxin-independent variability in transcript level, derives from at least these identifiable sources: (1) the large set of transcripts assayed (12 599) per chip and statistical fluctuation; (2) the heterogeneous nuclear backgrounds of the individual FRDA patient and control samples; (3) the different cell lineages used, which have a profound effect on gene expression; and (4) the experimental variation resulting from RNA extraction, cDNA reactions and cRNA labeling and other biochemical steps carried out on different days on different samples.

To trap or filter the frataxin-dependent signals, we used at the minimum a 3x3 study design in each cell type, i.e. three independent FRDA patient lymphoblasts, three control lymphoblasts, three independent FRDA fibroblasts, three control fibroblasts, three independent frataxin^RNA NT2 transfections, and three NT2 control microarrays. Thus, the mean frataxin-dependent alterations in gene expression had to be consistently altered at the P<0.05 level in three independent frataxin-deficient samples from the mean control values, to be included on the frataxin-dependent list within a cell type (i.e. Table 1).

The second trap or filter was cell autonomous gene expression, or CAGE. This analysis asks which of the significant frataxin-dependent changes within a cell type are significantly confirmed as frataxin-dependent in at least two cell types. This CAGE analysis asks in effect which frataxin-dependent alterations are strong enough to be observed independently of the huge effects of cell lineage on the transcriptome.

We can demonstrate with statistical significance that there are cell-autonomous frataxin-dependent alterations in transcripts in the following way. First, at the criteria of 1.5-fold change and P<0.05, the mean fraction of ‘frataxin-dependent’ genes among all three cell types (lymphoblasts, fibroblasts and NT2 cells) is 0.034, or 3.4%. Given this observation, we then generate our null hypothesis, i.e. that if there are no cell-autonomous, frataxin-dependent effects on gene expression, then we should observe that the number of frataxin-dependent genes shared between two cell types is the random product of these two frequencies, i.e. (0.034) (0.034)=0.001156/2=0.000578. (The 0.001156 is divided by 2 because our random expectation is that half of the changes are likely to be in the same direction in mutants versus controls.) This means that our random expectation of frataxin-dependent genes shared between any two cell types is (0.000578) (12 599 genes on the chip)=7.3 genes between two cell types. However our actual observed number of shared genes is 48 genes/3 comparisons=16 on average between two cell types, more than doubling our expectation of 7.3 genes, an increase which is significant at the P<0.01 level by a chi-squared test. Thus there is a doubling of the number of frataxin-dependent, cell-autonomous genes shared between cell types relative to the number one would expect by random chance. This excess both increases our confidence of the frataxin-dependence of the expression of these genes, and also this ratio of frataxin-dependent, cell-autonomous genes over the random expectation is an internal measure of the signal-to-noise within the experiment, i.e. a measure of frataxin-dependence on gene expression.

Thus, using only the original data quality criteria, and the additional CAGE criterion that genes must be expressed in the same direction in mutants versus controls in at least two cell types, these restrictions produced the list of 48 frataxin-dependent genes seen in Table 2.

View this table: [in this window] [in a new window]   Table 2. Cell-autonomous effects of frataxin deficiency

 
Of the five hypotheses for frataxin's function, there is multiple support only for deficiencies in the SAA/ISC pathway
Although there were alterations in transcripts whose gene products function in multiple biochemical pathways, there was not a single statistically significant and CAGE-verified ‘hit’ in any pathway relating to four of the hypotheses, i.e. iron transporter, iron storage, antioxidant and oxidative phosphorylation stimulator. By contrast, there were seven hits in the SAA and ISC biosynthetic pathways that are directly connected to each other, and we will refer to this contiguous pathway as SAA/ISC. These seven hits were in phosphoserine phosphatase, phosphoserine phosphatase-like, phosphoserine aminotransferase, seryl-tRNA synthetase, mitochondrial serine hydroxymethyltransferase, cystathionase and rhodanese. All seven of these transcripts were down-regulated (Table 2). The first four of these down-regulated transcripts are related to serine synthesis, which is condensed with homocysteine to make cystathionine, which is cleaved by cystathionase (also downregulated in mutants) to make cysteine, which is the sulfur donor for ISC synthesis, which is presumably catalyzed by the mammalian homologs of the bacterial ISC system, ISC-U and ISC-S (Fig. 2). The function of the mitochondrial protein rhodanese, which is also down-regulated in mutants, has been known for several years, i.e. to repair the iron–sulfur clusters in multiple mitochondrial ISC enzymes (26).

View larger version (28K): [in this window] [in a new window]   Figure 2. Down-regulated transcripts of frataxin-deficient cells map to the SAA pathway. Transcripts decreased in at least two frataxin-deficient cell types are highlighted in bold; in addition the cystathionine synthase transcript is decreased specifically in FRDA fibroblasts.

 
In addition to these seven ‘direct hits’ on the SAA/ISC pathway, we have also included two ‘indirect hits’ in Table 2, i.e. quinolinate phosphoribosyltransferase and eIF1A genes in the SAA/ISC category, for the following reasons. First, there is a very strong frataxin-dependent and cell-autonomous up-regulation in quinolinate phosphoribosyl-transferase (Table 2), an important enzyme in the synthesis of nicotinamide adenine dinucleotide, NAD. Studies in bacteria have demonstrated that a common consequence of ISC deficiency is a growth defect which is rescued by supplementing with the NAD precursor nicotinate (27), to which the up-regulation of the NAD biosynthetic enzyme quinolinate phosphoribosyl transferase is a possible response.

Secondly, the translation factor eIF1A is a remarkably conserved protein of translation initiation in multiple evolutionary kingdoms (28), and is responsible for recruiting the ‘initiator’ Met-tRNA to the AUG ‘start’ site on bacterial, mitochondrial and mammalian mRNAs. In bacteria and mitochondria this translation-initiating Met-tRNA is formylated. The formylation of this translation-initiation mitochondrial met-tRNA requires the substrate N¹⁰-formyl-THF, whose synthesis is dependent on the synthesis of N⁵,N¹⁰ methylene tetrahydrofolate, which is the product of the mitochondrial serine hydroxymethyltransferase. Thus a defect in mitochondrial serine hydroxymethyltransferase transcript that we have observed by microarray analysis should result in a decrease in folate metabolites, which should cause decreased concentration of the initiating formylmethionine–tRNA, which should decrease initiation of mitochondrial protein synthesis, to which the large induction of the eIF1A is a possible response, as the mitochondrion tries to drive the initiation of mitochondrial translation by increasing the eIF1a ‘partner’ in the ternary translation-initiation complex.

By contrast, we did not observe a significant and cell-autonomous change in these iron-related genes on the chip in comparisons between frataxin-deficient and control cells: iron responsive element 1, iron responsive element binding protein 2, ATX1 homolog 1, ferritin pseudogene 1, ferritin light polypeptide, ferritin heavy polypeptide, transferrin receptor, transferrin, lactotransferrin and hephaestin, lending no support to a role for iron-handling by frataxin. However, some genes involved in iron metabolism, including ABC7 and mtABC, were not included on this chip, so it is not possible to completely exclude a role of frataxin in iron handling, only to say that it appears much less likely than a primary role in ISC biosynthesis based on these data, with this selection of 12 599 genes represented on this Affymetrix U95Av2 chip.

We also did not observe significant and cell-autonomous differences in the oxidative stress genes mitochondrial superoxide dismutase, copper chaperone for superoxide dismutase, cytosolic superoxide dismutase extracellular superoxide dismutase and antioxidant protein 2, lending no support to a direct antioxidant role for frataxin.

There were only three mitochondrial genes consistently altered in the analysis: frataxin, mitochondrial serine hydroxymethyltransferase and rhodanese. The 50% down-regulation of frataxin observed in patient lymphoblasts and fibroblasts was consistent with our earlier estimates (15), and the 70% down-regulation in the RNAi NT2 model confirmed our quantitative RT–PCR results. One hypothesis proposed previously is that frataxin is a stimulator of mitochondrial oxidative phosphorylation. However, the down-regulation of serine hydroxymethyltransferase and rhodanese are each more consistent with the hypothesis of a defect in ISC/SAA in FRDA, rather than a generalized stimulator of oxidative phosphorylation.

Thus, of genes that could be imagined to participate in the pathways in which frataxin has been proposed to function, we observe seven clear hits in the linked SAA/ISC pathway, and none in any the other four hypotheses for frataxin's function, to the extent of resolution of the 12 599 genes on this chip and our study design.

Strong frataxin-dependent overexpression of multiple apoptosis transcripts
A second very strongly represented category that was obvious upon inspection of the 48 genes, was of nine apoptosis-related transcripts (Table 2), of which eight of the changes were increases. Many of these up-regulated apoptosis transcripts were in the TNF/INF/Fas dependent pathway. These data support a frataxin-dependent up-regulation of apoptosis transcripts. This may provide a molecular basis for our earlier observation that FRDA cells are sensitized to apoptotic stimuli, and our earlier biochemical observation that this increased apoptosis response is caspase-3 dependent and zvad-inhibitable (12,24).

Some frataxin-dependent genes are not related to SAA/ISC or apoptosis
In addition to the two major frataxin-dependent categories SAA/ISC and apoptosis discussed above, there were 29 other genes consistently altered in a frataxin-dependent way in more than one cell type. However none of these fit any of the other hypotheses, i.e. iron transporter, iron storage protein or mitochondrial oxidative phosphorylation stimulator. Some of these hits are likely to be the consequences of random transcriptional ‘noise’ as discussed above, and others are likely to be in fact frataxin-dependent cell-autonomous effects that are a downstream consequence of frataxin-deficiency.

Confirmation of microarray-detected alteration in SAA/ISC and apoptosis pathways by RT–PCR
Our quantitative real-time RT–PCR of the selected transcripts tumor necrosis factor and rhodanese strongly confirmed the accuracy of the Affymetrix microarray platform (data not shown). In addition, we investigated the expression of two recently cloned genes not represented on the chip that are the human homologs of the two major bacterial genes that participate in ISC biosynthesis, i.e. ISC-S and ISC-U, by RT–PCR. The transcript level of ISC-S was decreased in lymphoblasts, but the transcript level of ISC-U was not (data not shown).

Confirmation of decreased SAA concentration in frataxin-deficient cells
The down-regulation of seven SAA/ISC transcripts observed by microarray analysis is predicted to cause a decrease in the steady-state levels of the biochemical intermediates in this pathway. Thus we assayed steady-state levels of 23 amino acids in six sets of neural NT2 cells transfected with frataxin^RNAi, and also in mitochondria from FRDA and control lymphoblasts (Tables 3 and 4). In neural whole cell extracts, only four amino acids of the 23 were significantly (P<0.05) altered from control values in the transfected cells, and three of them were related to SAA metabolism: cystathionine, serine and homocystine. The mean values were a 35% decrease of cystathionine, a 19% decrease of serine and a 78% decrease of homocystine (Table 3). The level of cyst(e)ine in neural cells was extremely low, and below the sensitivity of the methods used to detect them.

View this table: [in this window] [in a new window]   Table 3. Cytoplasmic amino acid analysis

 

View this table: [in this window] [in a new window]   Table 4. Mitochondrial amino acid analysis

 
In mutant lymphoblast mitochondria, cyst(e)ine, the proximal donor of sulfur to iron–sulfur clusters, was significantly decreased, as were ornithine and glycine. The origin of mitochondrial cysteine is not currently known. Thus, the frataxin-dependent mitochondrial cysteine deficiency could be the result of either a defect of mitochondrial import of cytoplasmically-synthesised cysteine (or cysteinesulfinate), or could be the result of a defect of de novo mitochondrial cysteine biosynthesis. There was no decrease in mitochondrial methionine, consistent either with the idea of defective cysteine transport, or also with defective mitochondrial cysteine biosynthesis if mitochondrial cysteine is biosynthesized directly from serine and PAPS and sulfide as is the case in bacteria.

The decrease in mitochondrial glycine is likely to be a consequence of the decreased transcript level of mitochondrial serine hydroxymethyltransferase, which is required for the conversion of serine to glycine in mitochondria.

The mean ornithine concentration was significantly lower in mutants in both neural whole cell extracts and lymphoblast mitochondria, suggesting that this is a consistent consequence of frataxin deficiency (Tables 3 and 4). A recent microarray study has attempted to transcriptionally define the response to free sulfur, and has shown that increased free sulfur induces the bacterial/mitochondrial carbamoyl phosphate synthetase, also known as ornithine carbamoyltransferase (29). Thus, if frataxin deficiency results in defective mitochondrial ISC assembly or repair, and a consequent increase in free sulfur in mitochondria, this is by homology to the bacterial case predicted to increase the activity of the mitochondrial urea cycle enzyme carbamoyl phosphate synthetase, which should result in a lower steady-state concentration of mitochondrial and cytoplasmic ornithine. Consistent with the prediction of higher mitochondrial carbamoyl phosphate synthetase activity, we observe lower mean ammonia concentrations in mutant mitochondria.

A frataxin-specific defect in ISC-U expression in FRDA lymphoblasts
To further elucidate the specific effects of frataxin deficiency on the ISC biosynthetic pathway, the expression of the mammalian ISC-U and ISC-S homologs of the bacterial ISC biosynthetic pathway was investigated by western blot. ISC-U is the human presumptive iron–sulfur ‘scaffold’ protein by homology to the bacterial IscU gene, and the human ISC-S is the presumptive cysteine desulfurase by homology to bacterial IscS (Figs 2,3A and B). ISC-U protein level showed a clear and consistent decreased expression in the lymphoblast and fibroblast FRDA cells. This was interesting because there was not a large decrease in the steady-state level of the ISC-U transcript by RT–PCR in lymphoblasts, suggesting a post-transcriptional regulation. Co-immunoprecipitation studies suggested a weak interaction between frataxin and ISC-U, consistent with a post-transcriptional role on ISC-U stabilization (data not shown). By contrast, ISC-S protein levels were slightly decreased, consistent with their decreased transcript level by RT–PCR.

View larger version (60K): [in this window] [in a new window]   Figure 3. Protein expression of ISC-S and ISC-U in FRDA cells. (A) Protein expression of ISC-U and (B) ISC-S in FRDA cells was detected by western blot. Lane p, extract of RD4 cells as positive control; lane c1-2, two lymphoblast controls; lane p1–3, FRDA lymphoblasts; Fc, fibroblasts control; Fm, FRDA mutant fibroblast.

 
Confirmation of a frataxin-dependent ISC defect in mitochondrial ISC enzyme activity
To further confirm a frataxin-dependent defect in ISC biogenesis, we isolated mitochondria and measured the activities of the mitochondrial ISC-dependent enzymes aconitase and succinate dehydrogenase, which were deficient in the frataxin-deficient cells (data not shown), consistent with a defect in ISC biogenesis.

Increased peroxidative stress and decreased GSH/GSSG ratio, but no decreased GSH in frataxin-deficient cell extracts or mitochondria
One puzzling feature of FRDA is the sensitivity of mutant cells to oxidative stress (30–32). The microarray data above and the biochemical analysis strongly support a deficiency in the SAA/ISC pathway, from which cysteine is made. In turn glutathione, the major antioxidant of the cell, is made from cysteine. Thus one possible explanation for the oxidant sensitivity of FRDA cells is that frataxin-deficiency causes glutathione deficiency, and thus less antioxidant protection of the FRDA cells. An alternative explanation is that, as a result of the SAA/ISC defect, these frataxin-deficient mitochondria themselves produce more reactive oxygen species. To discriminate between these two possibilities, we confirmed that FRDA mitochondria produced higher burdens of peroxides, and showed that these higher levels of peroxides were rescued by frataxin transfection (Fig. 4A). By contrast, although oxidized glutathione (GSSG) levels are clearly and significantly elevated in frataxin-deficient cells (Fig. 4C and D), indicating increased incipient oxidative stress, we observed no significant decrease of total glutathione in frataxin-deficient cell extracts or mitochondrial extracts (Fig. 4B). These data are consistent with the idea that the oxidant-sensitivity of FRDA cells derives from a higher production of mitochondrial reactive oxygen species (ROS), rather than a deficiency in antioxidant protection, and is consistent with recent data from yeast suggesting that oxidant-sensitivity is secondary to an ISC defect (20).

View larger version (19K): [in this window] [in a new window]   Figure 4. (A) Mitochondrial peroxide production by the dichlorofluorescein diacetate assay in control mitochondria, FRDA mitochondria, and mitochondria from FRDA lymphoblasts transfected with frataxin (FRDA-t). (B–D), GSSG, GSH, and GSH/GSSG ratio in control lymphoblasts and FRDA lymphoblasts. The error bars represent standard deviation and *P<0.05 determined by Student's t-test (**FRDA-t compared with FRDA).

 
Partial rescue of the frataxin-dependent oxidant-sensitive phenotype using a sulfide compound
Given the multiple transcriptional and biochemical indications that a deficiency of sulfur chemistry is primary in frataxin-deficient human cells, we attempted to rescue the frataxin-deficient cells with various sulfur compounds and other compounds that would be predicted to supplement the SAA pathway. Using an assay that distinguishes between mutant and control cells based on oxidant sensitivity (14,31), we pre-treated the cells with several compounds to attempt to rescue them from the oxidant-sensitivity. The control cells were resistant to the oxidant challenge, as expected. Of multiple compounds that were tried, sodium sulfide, sodium sulfate, sodium sulfite, cystathionine, vitamin B₁₂, N-acetyl cysteine and others, only pre-treatment with sodium sulfide provided partial but significant rescue from the oxidant challenge (Fig. 5). This was interesting in multiple contexts, including that ISCs can form spontaneously, and that sulfide is a biologically active form of sulfur, and that a critical step in ISC synthesis is thought to be the transfer of persulfide sulfur from ISC-S to ISC-U (33). Also, if there is unique mitochondrial biosynthesis of cysteine, it is likely to be the ‘bacterial’ pathway, which requires only sulfide and serine. Thus sulfide might provide rescue either by facilitating sulfur transfer to ISCs or by supplementing mitochondrial cysteine levels.

View larger version (28K): [in this window] [in a new window]   Figure 5. Significant protection of FRDA cells from oxidative challenge by a specific sulfur compound. FRDA and control fibroblasts were grown in DMEM medium with or without 200 µM of sodium sulfide, sodium sulfate or sodium sulfite ,or 10 µM vitamin B12 for 3 h, and then cells were re-plated, and challenged with 200 µM t-butyl hydroperoxide (t-BOOH) for 24 h, and cell viability was determined. Error bars represent standard deviation and *P<0.05, determined by paired Student's t-test.

 

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIAL AND METHODS REFERENCES

 
Microarray analysis supports a role for frataxin in mitochondrial ISC/SAA biosynthesis
In the last 6 years, five functions have been proposed for frataxin's mitochondrial function: iron transporter, ISC biosynthetic protein, iron storage protein, antioxidant and OXPHOS stimulator. Most of the hypotheses of frataxin function were primarily derived from work in non-human models, which are significantly divergent at the genetic and biochemical levels from humans. Using stringent data quality criteria, and a 3x3 study design in each cell type line, we identified hundreds of genes that were nominally frataxin-dependent within each cell type. These were refined to 48 genes by further specification of the CAGE criterion. We have demonstrated that these 48 genes are in statistically significant excess of the 21.9 genes that were expected under a random model, if frataxin-dependent effects were not cell-autonomous.

Of the five current, existing hypotheses for frataxin function, the hypothesis of an alteration in ISC/SAA metabolism received seven direct hits, and two indirect associations, and the other four hypotheses each received zero (direct or indirect). Even if we make the most conservative test possible, i.e. consider all non-ISC hypotheses in one category, and only ISC in the other, and make the very conservative estimate that half (4.5) of the nine hits should be in each of these two categories, the observation of 9 : 0 is still a significant deviation from random expectation of 4.5 : 4.5 at the P<0.05 criterion.

None of the eight transcripts on the U95Av2 chip that are related primarily to iron transport or metabolism were significantly frataxin-dependent with respect to the CAGE criteria. This is consistent with our earlier observation of no significant increase in total mitochondrial iron concentration in human frataxin-deficient cells (14,15). Also, we observed no consistent alterations in antioxidant genes, which is consistent with our current results on glutathione levels, i.e. it appears that there is not a defect in the biosynthesis of glutathione. Regarding frataxin as a stimulator of oxidative phosphorylation, there were only two consistent changes in nucleus-encoded mitochondrially-localized genes besides frataxin, i.e. serine hydroxymethyltransferase and rhodanese, and their alterations are better explained by an SAA/ISC hypothesis, rather than any of the other four hypotheses. Thus the microarray data provide significant support only for an SAA/ISC hypothesis for FRDA, and no support for any of the other four hypotheses, i.e. iron transport, iron storage, antioxidant or mitochondrial stimulator. While we can certainly not exclude that frataxin participates in those functions, these microarray data do not support provide any support for them.

Multiple molecular and biochemical data confirm an alteration in ISC/SAA metabolism in frataxin-deficient cells
The microarray-based suggestion of an alteration in ISC/SAA metabolism in frataxin-deficient cells has been strongly and multiply confirmed at the molecular and biochemical level.

Firstly, the decreases in steady-state transcript level of multiple genes involved in ISC/SAA metabolism (rhodanese, ISC-S, ISC-U) was confirmed by Q-RTPCR. The cysteine desulfurase ISC-S is thought to be the major sulfur donor to the growing iron–sulfur cluster, and rhodanese is a mitochondrial protein that repairs damaged ISCs in mitochondrial enzymes (26), while ISC-U is thought to function as the scaffold on which ISCs are built (34). The fact that rhodanese transcripts are deficient in frataxin-deficient cells, combined with the recent data that frataxin-deficient yeast only have a partial ISC biosynthetic defect (19) suggests that defective ISC repair may be just as or more important as defective ISC biosynthesis in the pathophysiology of FRDA.

Secondly, the microarray-based prediction of a decreased flux of SAAs through their biosynthetic pathway was strongly confirmed by only finding significant decreases in SAAs themselves (cysteine, homocysteine, cystathionine), and in the SAA-related amino acids serine, glycine and ornithine. The deficiencies of cysteine, homocysteine and cystathionine are clearly directly in SAA and, similarly, serine is required for cystathionine synthesis, deficient mitochondrial glycine is probably the result of a deficient mitochondrial serine hydroxymethyltransferase activity, and free sulfur has been shown to stimulate bacterial carbamoyl phosphate synthetase, which should result in a decreased steady-state level of ornithine and mitochondrial ammonia, which were observed.

Thirdly, we isolated mitochondria and assayed the mitochondrial enzyme activities of aconitase and succinate dehydrogenase, two mitochondrial enzymes which require ISCs for their activity, and both of these activities were deficient in mutant cells. This is the most straightforward biochemical test for a deficiency of mitochondrial ISC biogenesis and/or maintenance.

Fourthly, the western data demonstrated a clear and consistent decrease in human ISC-U protein in FRDA cells, most pronounced in the FRDA lymphoblasts. This was particularly interesting in that we did not observe a difference in steady state ISC-U transcript level by quantitative RT–PCR, implicating a post-transcriptional mechanism of frataxin-dependent support of ISC-U expression. Studies elsewhere have identified a function for the bacterial HscB protein as a ‘co-chaperone’ of the bacterial Iscu scaffold to the HscA chaperone (L.Vickery, personal communication), and the HscB protein has appeared to co-evolve with frataxin-like proteins over evolutionary time (35).

Thus, one possible interpretation of the results is that in human cells frataxin has a HscB-like ‘co-chaperone’ function, and our co-immunoprecipitation data support a physical interaction between frataxin and ISC-U (data not shown). In the case of absence or deficiency of this function, ISC-U may exhibit defective folding, stability and/or transport, which may in turn result in a specific deficiency of the scaffold function in the ISC biogenesis process, which ultimately impairs ISC biogenesis and consequently rhodanese-mediated ISC repair as well. Deficiencies in the scaffold structure or function could be predicted to cause a premature release of free iron (and free sulfur), consistent with our previous studies, demonstrating an increased concentration of free iron in FRDA mitochondria (15). An increase in mitochondrial free iron is expected to cause increased Fenton Chemistry and increased reactive oxygen species (ROS) production, which we observe specifically in FRDA mitochondria, and which is rescued by frataxin. An increased ROS production by FRDA mitochondria is also consistent with the higher concentrations of GSSG we observe in mutant cells. Thus the sources of the increased mitochondrial ROS in FRDA mitochondria could include the increased free iron, malformed ISCs themselves, and the redox-active ISC-depleted apoproteins.

Relationship of the SAA defect to the ISC defect in human cells
The simplest explanation of the data appears to be that frataxin acts to stabilize ISC-U and thus to support ISC biosynthesis and also maintenance, and that a defect in this process is sensed by the cell, which then inhibits multiple transcripts along the SAA/ISC pathway through some sort of feedback process. This is the most consistent with all of our data and is the most direct.

However, another explanation of the data is that the mitochondrial ISC defect is an indirect consequence of a mitochondrial cysteine deficiency, which is itself the result of a frataxin-dependent step in either the mitochondrial-specific pathway of serine or cysteine import, or cysteine biosynthesis. If mitochondrial cysteine biosynthesis were like bacterial cysteine biosynthesis, only a very few substrates are required, i.e. serine and sulfide.

The least direct explanation of the data is that the upstream decrease in cytoplasmic serine-specific transcripts, and the decrease in mitochondrial cysteine levels downstream, are both the consequence of a defect in mitochondrial serine hydroxymethyltransferase. In this interpretation decreased mitochondrial serine hydroxymethyltranferase activity causes: (1) decreased mitochondrial serine utilization and a consequent feedback and inhibition of cytoplasmic serine synthesis; (2) a defect in mitochondrial cysteine synthesis and/or transport; and (3) a decrease in mitochondrial folate production which precipitates problems in initiation of protein synthesis and apoptosis. Note that in this third explanation ISC deficiency is a consequence of the frataxin-dependent defect, but is not necessarily the proximal cause of cell death.

Microarray data suggest multi-step consequences of frataxin deficiency
Consistent with our earlier observation of an increased apoptotic activity in FRDA cells at the cellular and biochemical levels (14,15), we observe in this molecular microarray study a strong up-regulation of multiple pro-apoptotic transcripts in frataxin-deficient cells such as caspase-1 and tumor necrosis factor. Multiple groups have demonstrated that pharmacological or genetic inhibition of translation is a extremely potent and rapid inducer of apoptosis (36,37). Thus, the inhibition of mitochondrial serine hydroxymethyltransferase is predicted to decrease concentrations of N⁵,N¹⁰ methylene tetrahydrofolate, from which N¹⁰-formyl tetrahydrofolate is made, which is a required substrate for the formylation of methionine tRNA, to which the increased eiF1A is a likely response, inhibiting protein synthesis, which has been demonstrated to induce apoptosis. These data indicate that one possible mechanism of FRDA is (-) frataxin(-) ISC(-) SAA(-) SHMTtranslational defectapoptosiscell death, or alternatively that (-) frataxin(-) SHMT(-) SAA & translational defectapoptosiscell death.

A multi-step FRDA mechanism may allow multiple points of intervention
If either of the stepwise mechanisms described in the previous section are correct, then they suggest multiple avenues for intervention into FRDA. For example, it has been known for some time that ISCs form spontaneously from iron and sulfur, and the discovery of a biochemical pathway for ISC synthesis was unexpected by many. Thus providing biologically available sulfur to FRDA patients may supplement the spontaneous biogenesis or maintenance of ISCs, or de novo mitochondrial cysteine biosynthesis if this occurs. Our data suggest that providing sulfur in its sulfide oxidation state is the only one that provides significant protection from the oxidant-sensitivity of FRDA cells. If the frataxin-dependent step is earlier in the ISC/SAA pathway, resulting in an ISC defect that is only the consequence of decreased mitochondrial SAA metabolites, then this could presumably be rescued by providing one or more of these SAA metabolites to mitochondria. The incipient oxidative stress which we have suggested may be the result of increased free mitochondrial iron could presumably be rescued by mitochondrially permeable iron chelators, and the use of antioxidant drugs. Lastly, the induction of apoptotic transcripts suggests that rescue with formylmethionine or specific inhibitors of apoptosis may suppress the apoptotic effects of frataxin deficiency. Thus these data suggest that there are multiple steps in the pathophysiological process downstream of frataxin deficiency, each of which could serve as independent routes to therapy.

	MATERIAL AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIAL AND METHODS REFERENCES

 
Human genome gene chips U95Av2 were purchased from Affymetrix (Santa Clara, CA, USA). The reagents and kits for whole microarray assays were purchased from companies recommended by the manual. The reagents for RT–PCR were purchased from Invitrogen (Gaithersburg, MD, USA).

Cell lines
Lymphocytes from three FRDA patients with severe clinical phenotypes and from three normal controls were grown as previously described (15). Two of three mutant lines were transfected by full-length frataxin cDNA construct and generated two higher frataxin expression lines (15). The genotypes of the two lines (p131, p585) were described previously (15), and the genotype for cell line p218 was (GAA)₆₀₀/W173G. The oligonucleotide primers used for sequencing and for long-range PCR were previously described (38). Frataxin mRNA expression levels were detected using quantitative RT–PCR described as below (Fig. 6A).

View larger version (14K): [in this window] [in a new window]   Figure 6. Frataxin mRNA expression levels in controls and Friedreich's ataxia (FRDA) cells. Frataxin mRNA levels were detected by using quantitative RT–PCR. The levels represented percent of averages of controls. (A) Lymphoblasts: Ctl (average of three control lines), p131, p585 and p218, three FRDA lines, and p131-t and p585-t, two frataxin-transfected lines. (B) Fibroblasts: Ctl (average of three controls), p13, p1037 and p1035, three FRDA lines. The error bars represent standard deviation from three repeated experiments.

 
Three FRDA fibroblasts and three controls were grown in the DMEM media described previously (14). The genotype was determined by long-range PCR as above. The genotype of p1037 was (GAA)₇₆₀/(GAA)₇₆₀, p1035 (GAA)₆₆₀/(GAA)₆₆₀, p13 (GAA)₅₃₀/(GAA)₈₃₀. Frataxin mRNA expression levels were detected by quantitative RT–PCR as described below (Fig. 6B).

NT2 cells were grown in DMEM/F12 supplemented with 10% FBS, 50 µg/ml uridine, 1 µM sodium pyruvate, penicillin and streptomycin. After two passages, cells were used for frataxin RNA interference.

Quantitative RT–PCR analysis
Total RNA was prepared using Qiagen Mini Kit (Qiagen, CA, USA), and RNA concentration was determined by UV spectrophotometry. Reverse transcription of 1 µg RNA was performed using an RT–PCR kit (Invitrogen, Gaithersburg, MD, USA), and reactions were performed in a 20 µl volume. Two microlitres cDNA from 20x dilution of RT reaction were used for PCR. Quantitative PCR standard curves were set up for frataxin, beta-actin, Rhodanese, which method was previously described (15). The PCR primers for frataxin and beta-actin were previously described (15). The primers of PCR for Rhodanese are: forward, GTG GAT GTT CCG TGT GTT TG; reverse, TGC GAG AGA TCC ACC TTC TT. The PCR reaction and normalization were previously described in detail (15).

RNA interference (RNAi) for inhibiting frataxin expression in neural cells
N-tera2/D1 (NT2)cells were grown in DMEM/F12 media and plated at 2x10⁵ cells/well in six-well plates. The cells were incubated in medium without antibiotics for 16 h and then replaced with OPTM 1 medium without serum for 1 h. A mixture of 200 nM siRNA (frataxin, AAC GUG GCC UCA ACC AGA UUU; or scrambled, CAG UCG CGU UUG CGA CUG GdTdT) and Oligofectamine (Invitrogen, Gaithersburg, MD, USA) was added to each well for 4 h, after which DMEM/F12 containing 30% serum was added. Cells were harvested 48 h later and total RNA and protein were extracted. Quantitative RT–PCR was used to confirm the frataxin mRNA level (see above), and western blot was used to confirm the frataxin protein level (see below). Samples were used for microarray analysis.

Western blot analysis
Cells were lysed in a lysis buffer (15) at 4°C for 30 min. Insoluble material was removed by centrifugation, and equal amounts of lysates (20 µg) were used for immunoblotting. Protein concentration was estimated using the Bradford protein assay system (Bio-Rad). Samples were resolved on a 15% SDS–polyacrylamide gel and then transferred to a PVDF membrane (Millipore, Bedford, MA, USA) by electroblotting. After blocking with 4% non-fat dry milk, the blot was incubated with anti-frataxin polyclonal antibody or anti-ISC-U or anti-ISC-S (antibody was kindly provided by Drs Tracey Ruoualt and Wing-Hang Tong) and was developed with AP-conjugated secondary antibodies using a chemiluminescent substrate.

Oligonucleotide microarray analysis
Total RNA extraction was performed using the RNeasy Mini Kit (Qiagen, Valencia, CA, USA). Double-stranded cDNA was synthesized from 10 µg of purified RNA using the SuperScript Choice System (Invitrogen, Gaithersburg, MD, USA) and T₇- (dt)₂₄ primer (Genset, La Jolla, CA, USA). The cRNA was prepared and biotin-labeled by the Enzo in vitro transcription kit (Enzo Biochemical) and labeled cRNA was purified using the Qiagen RNeasy mini cleanup protocol. Fragmentation of the cRNA was performed at 95°C for 35 min in fragmentation buffer (40 mM Tris acetate, pH 8.1, 100 mM potassium acetate, and 30 mM magnesium acetate). The quality of RNA and fragmented cRNA was checked by gel electrophoresis. Hybridization of the fragmented cRNA to the Hu95Av2 chip (Affymetrix, Santa Clara, CA, USA) and subsequent steps were performed by the Microarray Core Facility at UC Davis. Briefly, 15 µg of fragmented cRNA were hybridized to the chip for 16 h at 45°C. The gene chips were automatically washed and stained with streptavidin-phycoerythrin using a fluidics station. Finally, probe arrays were scanned on an Agilent Genearray Scanner G2500 (Agilent Tech., Palo Alto, CA, USA). Affymetrix Microarray Suite 5.0 was used for data analysis. Absolute analysis was performed, which calculates a variety of metrics using probe array hybridization intensities, that are measured by the scanner. Some are used for background and noise calculations; others are used to calculate the expression level of each transcript (average difference) and to determine the presence or absence of a transcript. The absolute analysis results of different experiments were scaled to an average intensity of 400 and analyzed independently.

DNA-chip data analysis
DNA-chip analyzer (Dchip) is a software package implementing a model-based expression analysis of high-density oligonucleotide DNA array data. By pooling information across multiple arrays, it is possible to assess the standard errors for the expression indexes. High-level analysis in this approach includes comparative analysis and hierarchical clustering (25). All data used in our analysis were derived from the Affymetrix Microarray Suite 5.0 software. The DNA-chip imported the raw .cel files into its program and carried out normalization and set up model-expression data.

The transcripts were filtered as a set, and the following criterion was used: a criterion of 0.1<standard deviation/mean<10; present call in arrays is equal or more than 40%; and expression level is ≥10 in 80% samples or more. Of the 12 599 genes, 6193 satisfied this filtering criterion in lymphoblasts, 6269 genes in fibroblasts, and 4440 genes in neural cells.

To compare the expression levels between the cells with or without frataxin deficiency, the comparison analysis was performed in lymphoblasts (6193 genes), fibroblasts (6269 genes) and neural cells (4440 genes). The criteria were set as follows: fold change was equal to or greater than 1.5; P-value was ≤0.05; and the intensity of |E-B|≥40 (E=experimental=intensity of FRDA samples; B=background=intensity of controls). In lymphoblasts, 645 of the 6193 genes satisfied this comparison criterion, and in fibroblasts, 576 of 6269 genes, and in neural cells, 185 of 4440 genes (Table 1).

To determine which changes were consistent crossing different cell types with frataxin deficiency, we re-organized .dcp files and performed combining comparisons using DNA-chip data analyzer. Owing to large expression difference of genes from cell type to cell type, data from lymphoblasts, fibroblasts and neural cells were, respectively, normalized within each cell type, and model-based expression values were obtained for each cell type, and a .dcp file was generated for each sample. All .dcp files from three cell types were re-loaded to Dchip Data Analyzer, and comparisons within cell type were carried out using the criteria above, and one further criterion—that the difference in expression between mutant and control cells must be in the same gene and in the same direction, i.e shared between at least two cell types—which winnowed the list down to the 48 genes shown in Table 2.

Amino acid analysis
Samples from neural cells with or without frataxin deficiency were suspended in PBS buffer, and the cells were lysed by four cycles of a freeze-thaw treatment. After centrifugation of 5000g, supernatants were sampled for amino acid analysis using the Beckman 6300 amino acid analyzer (Beckman, Germany). The mitochondria were isolated from FRDA and control lymphoblasts as previously described (24), and amino acid analyses were done as above. The determination of protein concentration in the supernatant fraction was described as above. Most amino acids were steadily detected in all samples by using this analysis. The content of each amino acid presented as nmol/mg protein. All experiments were repeated at least three times.

Plating efficiency of FRDA fibroblasts
We tested the plating efficiency of FRDA fibroblasts exposed to 150–250 µM tBH tert-butyl hydroperoxide (tBOOH) as in Fujita et al. (39); the viability of FRDA cells was consistently lower than control fibroblasts. Cells were plated at 5x10⁵ cells/T25 cm² flask and were pre-treated with or without rescue reagents for 4 h, and then re-plated cells at the same density and treated with or without 150–250 µM tBH for 24 h. The viability of cells was determined by Trypan blue exclusion assay as previously described (14).

Total reduced glutathione (GSH) and oxidized (GSSG) amount
To assay total GSH amount, 4x10⁷ cells were collected and treated with 1 ml of 6% metaphosphoric acid (MPA) for 10 min at 0°C. The cell lysate was centrifuged at 12 000g for 10 min at 4°C. The supernatant was harvested, stored at 0°C and immediately used for total GSH-GSSG content assay described previously (40,41). Briefly, the sample for the assay was diluted 1 : 6–1 : 8 with 6% MPA. A 0.1 ml aliquot of soluble fraction were added in 1 ml containing 0.1 M KH₂PO₄+5 mM EDTA (pH=7.4), DTNB 0.5 mM, NADPH 0.8 mM and 2 units of GSH-reductase (Sigma). The reaction was carried out for 3 min and the activity measured at a =412 nm. Results were compared with a standard curve of GSH (0.1, 0.2 and 0.4 µg/ml) in 6% MPA and expressed as nmol/mg protein. To assay the amount of GSSG, the soluble fraction was treated for 1 h at room temperature with triethanolamine (6 µl/100 µl sample) and 2-vinylpyridine (2 µl/100 µl sample), an alkylating agent specific for reduced sulfhydryl groups. The assay was carried on as previously described. Results were compared with a standard curve of GSSG (0.05, 0.1 and 0.2 µg/ml) in 6% MPA and expressed as nmol/mg protein.

	ACKNOWLEDGEMENTS

 
We thank A. Wong, S. Danielson, and Q. Rogers for helpful comments, M. George and M. Rolston for microarray help; this work is supported by USPHS grants AG11967, AG16719, EY12245, and a pilot project supported by P30ES05707 to G.A.C.; and Telethon-Italia support to F.T. (grant no. E.0514) is also gratefully acknowledged.

	FOOTNOTES

 
^* To whom correspondence should be addressed. Tel: +1 5303046810; Fax: +1 5307549342; Email: gacortopassi{at}ucdavis.edu

	REFERENCES

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIAL AND METHODS REFERENCES

 

1. Durr, A., Cossee, M., Agid, Y., Campuzano, V., Mignard, C., Penet, C., Mandel, J.L., Brice, A. and Koenig, M. (1996) Clinical and genetic abnormalities in patients with Friedreich's ataxia. New Engl. J. Med., 335, 1169–1175.[Abstract/Free Full Text]

2. 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 an intronic GAA triplet repeat expansion. Science, 271, 1423–1427.[Abstract]

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

4. Foury, F. (1999) Low iron concentration and aconitase deficiency in a yeast frataxin homologue deficient strain. FEBS Lett., 456, 281–284.[CrossRef][Web of Science][Medline]

5. 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 yeast frataxin: implications for Friedreich ataxia. Am. J. Hum. Genet., 67, 549–562.[CrossRef][Web of Science][Medline]

6. Ristow, M., Pfister, M.F., Yee, A.J., Schubert, M., Michael, L., Zhang, C.Y., Ueki, K., Michael, M.D., 2nd, Lowell, B.B. and Kahn, C.R. (2000) Frataxin activates mitochondrial energy conversion and oxidative phosphorylation. Proc. Natl Acad. Sci. USA, 97, 12239–12243.[Abstract/Free Full Text]

7. 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.[CrossRef][Web of Science][Medline]

8. Gakh, O., Adamec, J., Gacy, A.M., Twesten, R.D., Owen, W.G. and Isaya, G. (2002) Physical evidence that yeast frataxin is an iron storage protein. Biochemistry, 41, 6798–67804.[CrossRef][Medline]

9. Foury, F. and Cazzalini, O. (1997) Deletion of the yeast homologue of the human gene associated with Friedreich's ataxia elicits iron accumulation in mitochondria. FEBS Lett., 411, 373–377.[CrossRef][Web of Science][Medline]

10. Bradley, J.L., Blake, J.C., Chamberlain, S., Thomas, P.K., Cooper, J.M. and Schapira, A.H. (2000) Clinical, biochemical and molecular genetic correlations in Friedreich's ataxia. Hum. Mol. Genet., 9, 275–282.[Abstract/Free Full Text]

11. Lamarche, J.B., Cote, M. and Lemieux, B. (1980) The cardiomyopathy of Friedreich's ataxia morphological observations in 3 cases. Can. J. Neurol. Sci., 7, 389–396.[Web of Science][Medline]

12. Cossee, M., Puccio, H., Gansmuller, A., Koutnikova, H., Dierich, A., LeMeur, M., Fischbeck, K., Dolle, P. and Koenig, M. (2000) Inactivation of the Friedreich ataxia mouse gene leads to early embryonic lethality without iron accumulation. Hum. Mol. Genet., 9, 1219–1226.[Abstract/Free Full Text]

13. Delatycki, M.B., Paris, D.B., Gardner, R.J., Nicholson, G.A., Nassif, N., Storey, E., MacMillan, J.C., Collins, V., Williamson, R. and Forrest, S.M. (1999) Clinical and genetic study of Friedreich ataxia in an Australian population. Am. J. Med. Genet., 87, 168–174.[CrossRef][Web of Science][Medline]

14. Wong, A., Yang, J., Cavadini, P., Gellera, C., Lonnerdal, B., Taroni, F. and Cortopassi, G. (1999) The Friedreich's ataxia mutation confers cellular sensitivity to oxidant stress which is rescued by chelators of iron and calcium and inhibitors of apoptosis. Hum. Mol. Genet., 8, 425–430.[Abstract/Free Full Text]

15. Tan, G., Chen, L.S., Lonnerdal, B., Gellera, C., Taroni, F.A. and Cortopassi, G.A. (2001) Frataxin expression rescues mitochondrial dysfunctions in FRDA cells. Hum. Mol. Genet., 10, 2099–2107.[Abstract/Free Full Text]

16. Puccio, H. and Koenig, M. (2000) Recent advances in the molecular pathogenesis of Friedreich ataxia. Hum. Mol. Genet., 9, 887–892.[Abstract/Free Full Text]

17. Garland, S.A., Hoff, K., Vickery, L.E. and Culotta, V.C. (1999) Saccharomyces cerevisiae ISU1 and ISU2: members of a well-conserved gene family for iron-sulfur cluster assembly. J. Mol. Biol., 294, 897–907.[CrossRef][Web of Science][Medline]

18. Schilke, B., Voisine, C., Beinert, H. and Craig, E. (1999) Evidence for a conserved system for iron metabolism in the mitochondria of Saccharomyces cerevisiae. Proc. Natl Acad. Sci. USA, 96, 10206–10211.[Abstract/Free Full Text]

19. 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.[Abstract/Free Full Text]

20. Muhlenhoff, U., Richhardt, N., Ristow, M., Kispal, G. and Lill, R. (2002) The yeast frataxin homolog Yfh1p plays a specific role in the maturation of cellular Fe/S proteins. Hum. Mol. Genet., 11, 2025–2036.[Abstract/Free Full Text]

21. Lutz, T., Westermann, B., Neupert, W. and Herrmann, J.M. (2001) The mitochondrial proteins Ssq1 and Jac1 are required for the assembly of iron sulfur clusters in mitochondria. J. Mol. Biol., 307, 815–825.[CrossRef][Web of Science][Medline]

22. Gakh, O., Adamec, J., Gacy, A.M., Twesten, R.D., Owen, W.G. and Isaya, G. (2002) Physical evidence that yeast frataxin is an iron storage protein. Biochemistry, 41, 6798–6804.[CrossRef][Medline]

23. Schulz, J.B., Dehmer, T., Schols, L., Mende, H., Hardt, C., Vorgerd, M., Burk, K., Matson, W., Dichgans, J., Beal, M.F. et al. (2000) Oxidative stress in patients with Friedreich ataxia. Neurology, 55, 1719–1721.[Abstract/Free Full Text]

24. Elbashir, S.M., Harborth, J., Lendeckel, W., Yalcin, A., Weber, K. and Tuschl, T. (2001) Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature, 411, 494–498.[CrossRef][Medline]

25. Li, C. and Wong, W.H. (2001) Model-based analysis of oligonucleotide arrays: expression index computation and outlier detection. Proc. Natl Acad. Sci. USA, 98, 31–36.[Abstract/Free Full Text]

26. Bonomi, F., Pagani, S. and Kurtz, D.M., Jr. (1985) Enzymic synthesis of the 4Fe-4S clusters of Clostridium pasteurianum ferredoxin. Eur. J. Biochem., 148, 67–73.[Web of Science][Medline]

27. Schwartz, C.J., Djaman, O., Imlay, J.A. and Kiley, P.J. (2000) The cysteine desulfurase, IscS, has a major role in in vivo Fe-S cluster formation in Escherichia coli. Proc. Natl Acad. Sci. USA, 97, 9009–9014.[Abstract/Free Full Text]

28. Sorensen, H.P., Hedegaard, J., Sperling-Petersen, H.U. and Mortensen, K.K. (2001) Remarkable conservation of translation initiation factors: IF1/eIF1A and IF2/eIF5B are universally distributed phylogenetic markers. IUBMB Life, 51, 321–327.[Web of Science][Medline]

29. Schut, G.J., Zhou, J. and Adams, M.W. (2001) DNA microarray analysis of the hyperthermophilic archaeon Pyrococcus furiosus: evidence for anNew type of sulfur-reducing enzyme complex. J. Bacteriol., 183, 7027–7036.[Abstract/Free Full Text]

30. Wong, A., Yang, J., Danielson, S., Gellera, C., Taroni, F. and Cortopassi, G. (2000) Sensitivity of FRDA lymphoblasts to salts of transition metal ions. Antioxid. Redox Signal., 2, 461–465.[CrossRef][Medline]

31. Chantrel-Groussard, K., Geromel, V., Puccio, H., Koenig, M., Munnich, A., Rotig, A. and Rustin, P. (2001) Disabled early recruitment of antioxidant defenses in Friedreich's ataxia. Hum. Mol. Genet., 10, 2061–2067.[Abstract/Free Full Text]

32. Shoichet, S.A., Baumer, A.T., Stamenkovic, D., Sauer, H., Pfeiffer, A.F., Kahn, C.R., Muller-Wieland, D., Richter, C. and Ristow, M. (2002) Frataxin promotes antioxidant defense in a thiol-dependent manner resulting in diminished malignant transformation in vitro. Hum. Mol. Genet., 11, 815–821.[Abstract/Free Full Text]

33. Urbina, H.D., Silberg, J.J., Hoff, K.G. and Vickery, L.E. (2001) Transfer of sulfur from IscS to IscU during Fe/S cluster assembly. J. Biol. Chem., 276, 44521–44526.[Abstract/Free Full Text]

34. Yuvaniyama, P., Agar, J.N., Cash, V.L., Johnson, M.K. and Dean, D.R. (2000) NifS-directed assembly of a transient [2Fe–2S] cluster within the NifU protein. Proc. Natl Acad. Sci. USA, 97, 599–604.[Abstract/Free Full Text]

35. Huynen, M.A., Snel, B., Bork, P. and Gibson, T.J. (2001) The phylogenetic distribution of frataxin indicates a role in iron-sulfur cluster protein assembly. Hum. Mol. Genet., 10, 2463–2468.[Abstract/Free Full Text]

36. Herbert, T.P., Fahraeus, R., Prescott, A., Lane, D.P. and Proud, C.G. (2000) Rapid induction of apoptosis mediated by peptides that bind initiation factor eIF4E. Curr. Biol., 10, 793–796.[CrossRef][Web of Science][Medline]

37. Polunovsky, V.A., Wendt, C.H., Ingbar, D.H., Peterson, M.S. and Bitterman, P.B. (1994) Induction of endothelial cell apoptosis by TNF alpha: modulation by inhibitors of protein synthesis. Exp. Cell Res., 214, 584–594.[CrossRef][Web of Science][Medline]

38. Gellera, C., Pareyson, D., Castellotti, B., Mazzucchelli, F., Zappacosta, B., Pandolfo, M. and Di Donato, S. (1997) Very late onset Friedreich's ataxia without cardiomyopathy is associated with limited GAA expansion in the X25 gene. Neurology, 49, 1153–1155.[Abstract/Free Full Text]

39. Fujita, R., Sirugo, G., Duclos, F., Abderrahim, H., Le Paslier, D., Cohen, D., Brownstein, B.H., Schlessinger, D., Mandel, J.L. and Koenig, M. (1992) A 530 kb YAC contig tightly linked to the Friedreich ataxia locus contains five CpG clusters and a new highly polymorphic microsatellite. Hum. Genet., 89, 531–538.[Web of Science][Medline]

40. Anderson, M.E. (1985) Determination of glutathione and glutathione disulfide in biological samples. Meth. Enzymol., 113, 548–555.[Web of Science][Medline]

41. Floreani, M., Petrone, M., Debetto, P. and Palatini, P. (1997) A comparison between different methods for the determination of reduced and oxidized glutathione in mammalian tissues. Free Radical Res., 26, 449–455.[Web of Science][Medline]

CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				F. Fortuna, P. Barboni, R. Liguori, M. L. Valentino, G. Savini, C. Gellera, C. Mariotti, G. Rizzo, C. Tonon, D. Manners, et al. Visual system involvement in patients with Friedreich's ataxia Brain, January 1, 2009; 132(1): 116 - 123. [Abstract] [Full Text] [PDF]

				E. Soragni, D. Herman, S. Y. R. Dent, J. M. Gottesfeld, R. D. Wells, and M. Napierala Long intronic GAA*TTC repeats induce epigenetic changes and reporter gene silencing in a molecular model of Friedreich ataxia Nucleic Acids Res., November 1, 2008; 36(19): 6056 - 6065. [Abstract] [Full Text] [PDF]

				A. Martelli, M. Wattenhofer-Donze, S. Schmucker, S. Bouvet, L. Reutenauer, and H. Puccio Frataxin is essential for extramitochondrial Fe S cluster proteins in mammalian tissues Hum. Mol. Genet., November 15, 2007; 16(22): 2651 - 2658. [Abstract] [Full Text] [PDF]

				Y. Shan, E. Napoli, and G. Cortopassi Mitochondrial frataxin interacts with ISD11 of the NFS1/ISCU complex and multiple mitochondrial chaperones Hum. Mol. Genet., April 15, 2007; 16(8): 929 - 941. [Abstract] [Full Text] [PDF]

				J. V. Llorens, J. A. Navarro, M. J. Martinez-Sebastian, M. K. Baylies, S. Schneuwly, J. A. Botella, and M. D. Molto Causative role of oxidative stress in a Drosophila model of Friedreich ataxia FASEB J, February 1, 2007; 21(2): 333 - 344. [Abstract] [Full Text] [PDF]

				I. Condo, N. Ventura, F. Malisan, B. Tomassini, and R. Testi A Pool of Extramitochondrial Frataxin That Promotes Cell Survival J. Biol. Chem., June 16, 2006; 281(24): 16750 - 16756. [Abstract] [Full Text] [PDF]

				V. Irazusta, E. Cabiscol, G. Reverter-Branchat, J. Ros, and J. Tamarit Manganese Is the Link between Frataxin and Iron-Sulfur Deficiency in the Yeast Model of Friedreich Ataxia J. Biol. Chem., May 5, 2006; 281(18): 12227 - 12232. [Abstract] [Full Text] [PDF]

				R. A. Schoenfeld, E. Napoli, A. Wong, S. Zhan, L. Reutenauer, D. Morin, A. R. Buckpitt, F. Taroni, B. Lonnerdal, M. Ristow, et al. Frataxin deficiency alters heme pathway transcripts and decreases mitochondrial heme metabolites in mammalian cells Hum. Mol. Genet., December 15, 2005; 14(24): 3787 - 3799. [Abstract] [Full Text] [PDF]

				B. Sturm, U. Bistrich, M. Schranzhofer, J. P. Sarsero, U. Rauen, B. Scheiber-Mojdehkar, H. de Groot, P. Ioannou, and F. Petrat Friedreich's Ataxia, No Changes in Mitochondrial Labile Iron in Human Lymphoblasts and Fibroblasts: A DECREASE IN ANTIOXIDATIVE CAPACITY? J. Biol. Chem., February 25, 2005; 280(8): 6701 - 6708. [Abstract] [Full Text] [PDF]

				H. Seznec, D. Simon, C. Bouton, L. Reutenauer, A. Hertzog, P. Golik, V. Procaccio, M. Patel, J.-C. Drapier, M. Koenig, et al. Friedreich ataxia: the oxidative stress paradox Hum. Mol. Genet., February 15, 2005; 14(4): 463 - 474. [Abstract] [Full Text] [PDF]

				C. M. Everett and N. W. Wood Trinucleotide repeats and neurodegenerative disease Brain, November 1, 2004; 127(11): 2385 - 2405. [Abstract] [Full Text] [PDF]

				D. Simon, H. Seznec, A. Gansmuller, N. Carelle, P. Weber, D. Metzger, P. Rustin, M. Koenig, and H. Puccio Friedreich Ataxia Mouse Models with Progressive Cerebellar and Sensory Ataxia Reveal Autophagic Neurodegeneration in Dorsal Root Ganglia J. Neurosci., February 25, 2004; 24(8): 1987 - 1995. [Abstract] [Full Text] [PDF]

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (30) Request Permissions Google Scholar Articles by Tan, G. Articles by Cortopassi, G. Search for Related Content PubMed PubMed Citation Articles by Tan, G. Articles by Cortopassi, G. Social Bookmarking          What's this?

Online ISSN 1460-2083 - Print ISSN 0964-6906

Copyright © 2009 Oxford University Press

Oxford Journals Oxford University Press

• Site Map
• Privacy Policy
• Frequently Asked Questions

Other Oxford University Press sites: Oxford University Press Oxford Journals China Oxford Journals Japan American National Biography Booksellers' Information Service Children's Fiction and Poetry Children's Reference Corporate & Special Sales Dictionaries Dictionary of National Biography Digital Reference English Language Teaching Higher Education Textbooks Humanities International Education Unit Law Medicine Music Online Products Oxford English Dictionary Reference Rights and Permissions Science School Books Social Sciences Very Short Introductions World's Classics