Search tips
Search criteria 


Logo of capmcAbout manuscripts / A propos des manuscritsSubmit manuscript / soumettre un manuscrit
Neurobiol Dis. Author manuscript; available in PMC 2010 June 15.
Published in final edited form as:
PMCID: PMC2886035

Gene expression profiling in frataxin deficient mice: Microarray evidence for significant expression changes without detectable neurodegeneration


Friedreich’s ataxia (FRDA) is caused by reduction of frataxin levels to 5–35%. To better understand the biochemical sequelae of frataxin reduction, in absence of the confounding effects of neurodegeneration, we studied the gene expression profile of a mouse model expressing 25–36% of the normal frataxin levels, and not showing a detectable phenotype or neurodegenerative features. Despite having no overt phenotype, a clear microarray gene expression phenotype was observed. This phenotype followed the known regional susceptibility in this disease, most changes occurring in the spinal cord. Additionally, gene ontology analysis identified a clear mitochondrial component, consistent with previous findings. We were able to confirm a subset of changes in fibroblast cell lines from patients. The identification of a core set of genes changing early in the FRDA pathogenesis can be a useful tool in both clarifying the disease process and in evaluating new therapeutic strategies.

Keywords: Friedreich’s ataxia, Microarray, Mouse model, Frataxin, Neurodegeneration, Knockin/knockout


Friedreich ataxia (FRDA), the most prevalent inherited ataxia, is most frequently caused by a GAA triplet repeat expansion within the first intron of the gene encoding for frataxin, a nuclear encoded mitochondrial protein (Campuzano et al., 1996). The mutation affects frataxin transcription, leading to severe reduction of protein levels in homozygous patients (Bidichandani et al., 1998). The normal function of frataxin, and how its deficiency ultimately leads to neuronal dysfunction and death, is not well understood. Deficiency of the yeast frataxin homolog protein Yfh1p causes a strong reduction in the assembly of mitochondrial proteins containing iron–sulfur clusters (ISC) (Muhlenhoff et al., 2002), and frataxin is required for ISC assembly in yeast mitochondria (Gerber et al., 2003; Lutz et al., 2001). These data support a specific role for frataxin in the biosynthesis of cellular ISC proteins, which may be in turn related to oxidative stress sensitivity and iron homeostasis alterations (Puccio and Koenig, 2002).

FRDA has been a challenging disease to model in mice. Homozygous deletion of frataxin in the mouse causes embryonic lethality a few days after implantation, demonstrating a pivotal role for frataxin during early development (Cossee et al., 2000). Heterozygous knockout mice show reduced (50%) frataxin levels, no obvious phenotype, and sporadic heart iron deposits after dietary iron load (Santos et al., 2003). Through a conditional gene-targeting approach, neuronal, cardiac (Puccio et al., 2001), and pancreatic (Ristow et al., 2003) frataxin knockout mice have been generated. These models show cardiac hypertrophy, large sensory neuron dysfunction, deficient ISC protein activities (Puccio et al., 2001), and diabetes due to reactive oxygen species increase, growth arrest, and apoptosis in pancreatic beta cells (Ristow et al., 2003).

However, in FRDA patients (Pianese et al., 2004) and in lymphoblastoid cell lines derived from FRDA patients (Campuzano et al., 1997) a residual frataxin activity (5–35% of normal levels) is present. Thus, animal models with FRDA reduction, rather than its complete absence, would be valuable to further explore the effects of moderate frataxin deficiency on cellular and organismal functioning. One such model has involved the generation of a mouse expressing frataxin only from a human transgene containing a small repeat expansion (Miranda et al., 2002). The presence of a homozygous (GAA)230 repeat expansion in frda mouse gene led to a reduction of frataxin levels to about 75% of the wild type (WT). After crossing this knockin mouse with a frataxin knockout, the resulting knockin/knockout offspring (KIKO) expressed 25–36% of the WT levels. These mice – when examined at 12 months of age – showed no obvious phenotype, no iron deposits, and no differences with controls after dietary and parenteral iron load (Miranda et al., 2002). This model therefore provides a significant advantage for gene expression studies aimed at understanding the consequences of frataxin deficiency, since it is not confounded by factors that often accompany but may not initiate neurodegeneration, such as cell loss or inflammation (Geschwind, 2000).

FRDA presents a striking regional distribution of neuropathological abnormalities, with constant involvement of cervical spinal cord, neuronal loss in brainstem nuclei, and fairly common loss of Purkinje cells in the cerebellar cortex (Lamarche et al., 1984). To address the FRDA regionality, we studied several brain regions from KIKO mice using DNA microarrays. We hypothesized that, at a time prior to any evidence of neurodegeneration, this would allow us to assess early cellular changes in tissues that were frataxin deficient, in the absence of detectable cell loss. Similar approaches have been used to assess biochemical changes prior to the onset of overt disease in other models of neurodegenerative conditions, such as spinocerebellar ataxia (SCA) 1 (Serra et al., 2004), ataxia with vitamin E deficiency (AVED) (Gohil et al., 2003), amyotrophic lateral sclerosis (Yoshihara et al., 2002), Huntington’s disease (Sipione et al., 2002), and in heterozygous carriers of ataxia telangiectasia (Watts et al., 2002). This approach has allowed us to gain insight into early molecular dysfunction caused by reduced frataxin levels and complements other recent studies in this area by highlighting key pathways for therapeutic intervention.

Materials and methods


Frataxin heterozygous knockout mice (frda+/−) were crossed with frda+/230GAA mice, to generate frataxin knockout/knockin mice (frda−/230GAA) and the offspring was genotyped as described (Miranda et al., 2002). In this study, four 6-month-old KIKO mice were compared to age and gender matched WT littermates. Total RNA from three brain regions, cervical spinal cord (SC), cerebellum (CB), and brainstem (BS), was extracted by acid phenol extraction (Trizol, GIBCO/BRL) as recommended by the manufacturer. The purity and quality of the extracted RNA were assayed by measuring the optical density at 260 and 280 nm (NanoDrop ND-100 Spectrophotometer, NanoDrop Technologies) and by gel electrophoresis on RNA assay chips (Agilent 2100 Bioanalyzer, Agilent Technologies). Four WT (two males and two females) and four KIKO (two males and two females) mice were compared. RNAs from WT samples of the same gender were pooled, and co-hybridized with KIKO samples (Fig. 1).

Fig. 1
Study design schematic. Four 6-month-old KIKO mice (2 males, 2 females) were compared to age and gender matched WT littermates. RNA extracted from each of three CNS regions was co-hybridized on microarray slides (n = 12, 4 for spinal cord, 4 brainstem, ...

Probe synthesis and hybridization

Labeled cDNA synthesis, hybridization, and signal detection were performed using the tyramide signal amplification (TSA, PerkinElmer) kit, according to the manufacturer’s protocols with minor modifications (Karsten et al., 2002). Briefly, two total RNA samples (1.5 μg) were reverse transcribed to fluorescein- and biotin-labeled cDNA, and hybridized on mouse 9K cDNA arrays (UCLA Microarray Core Facility,, including 9,150 genes and expressed sequence tags. This cDNA array based on the Incyte Unigem 1 mouse clone set was chosen because it had previously shown highly reproducible hybridizations (Karsten and Geschwind, 2002; Karsten et al., 2002). Probe signals were generated using Cy3 and Cy5 reporters, and the hybridization was duplicated with dye swapping, in order to eliminate the influence of dye bias effects (Liang et al., 2003; Yang et al., 2002). Eight hybridizations using 4 independent pairs were performed for each of the three brain regions, for a total of 24 microarray hybridizations. Two additional microarrays were used for homotypic control/control hybridizations.

Scanning and data analysis

Slides were scanned by the GMS 418 Array Scanner (Genetic Microsystems), and the resulting images were analyzed by ImaGene 4.2 (Biodiscovery) using auto segmentation measurements set 3 pixel buffer and width for background. Signals from the poor quality spots flagged by ImaGene software were ignored. The ImaGene-generated data were loaded onto Gene-Spring 6.0 (Silicon Genetics), the local background was subtracted, and only the signal intensities greater than the background were subjected to lowess normalization, to obtain intensity-dependent normalized ratios of KIKO to WT. After averaging the dye-swapped ratio pairs, every probe had at maximum 4 ratios for each of the three brain regions. EntrezGene (, Ensembl (, and GeneOntology ( were used to obtain nomenclature, sequence, and gene ontology (GO) information.

Statistical analysis

After ruling out the signal outliers, genes with at least 8 ratio measurements were analyzed by the one-sample Student’s t test, to select those differentially expressed across all the examined regions. In addition, one-way ANOVA with post hoc Tukey test was used to select those with specific regional changes. By means of EASE software (Hosack et al., 2003) differentially expressed genes with GO data available were searched for over-represented classes. EASE calculates the over-representation (within the subset of differentially expressed genes) of each GO functional cluster, with respect to the total number of genes assayed and annotated within each functional cluster.

Human cell lines

Primary fibroblast cell lines from 3 patients and 3 controls were obtained from Coriell Cell Repositories (Camden, NJ) and cultured in F12 Dulbecco’s modified essential medium with HEPES and glutamine (F12-DMEM, Invitrogen), with 10% calf serum and 1% penicillin–streptomycin. All cell lines were cultured at 37°C, in a humidified atmosphere of 5% CO2, 95% air.

Real-time quantitative PCR

Selected differentially expressed genes were assayed using real-time quantitative PCR (qRT-PCR) using SYBR green I as fluorescent dye. Total RNA (2 μg) from different animals (distinct from those studied in the microarray analysis) was treated with DNAse I (Promega) and converted into cDNA by SuperScript II kit (Invitrogen). The reactions were performed with 2× SYBR-green PCR Master Mix (BioRad), in a 25 μl volume. Assays were performed in triplicate, and analyzed using an ABI 7700 instrument (Applied Biosystems). The fold change was calculated using both standard curve analysis and the Pfaffl method (Pfaffl et al., 2002), using Hprt as reference gene. For each gene, data from at least 3 KIKO/WT pairs were averaged.


Genes differentially expressed across all the regions

We first identified genes that were differentially expressed across all the examined brain regions. Student’s t test analysis identified 185 sequences across 12 independent experiments that were significantly differentially expressed between mutant and WT animals. Among these genes, 116 were upregulated and 69 downregulated (Fig. 2). A list of selected differentially expressed sequences according to their proposed biological function is reported in Table 1. The observed changes were small, but statistically significant, with most genes with a fold change between 1.2 (0.2) and 1.5 (0.5), consistent with the little or absent phenotype.

Fig. 2
Differentially expressed genes in three CNS regions in frataxin deficient mice. 185 genes were identified as differentially expressed across all regions. An additional 105 genes showed expression changes with a regional distribution, following the gradient ...
Table 1
Differentially expressed genes between KIKO and control mice

Genes showing region-specific changes

Since FRDA involves degeneration of specific brain regions, rather than global neurodegeneration, it was also interesting to assess regional distinctions between mutant and WT mice. So, we next determined whether any genes were differentially expressed in some brain regions and not in others, identifying potential pathways that could underlie selective compensation or vulnerability. After ANOVA analysis, an additional 105 genes demonstrated a significant regional pattern of differential expression. Strikingly, the cervical spinal cord, which is the region most affected in the human disease, showed the most changes in gene expression (n = 61), followed by brainstem (n = 27), and cerebellum (n = 17). Moreover, these SC changes were biased to involve more downregulation than upregulation (Fig. 2, Table 1).

Functional categorization of gene expression changes

EASE analysis was used to help annotate genes relative to relevant functional categories. The subset of overall differentially expressed genes (including the genes showing regional changes) was classified according to GO biological process, cellular component, and molecular function (Table 2). Interestingly, the mitochondrial cellular component (along with ribonucleoprotein complex) showed a significant over-representation in this subset of genes, consistent with the mitochondrial localization of the frataxin protein across species, and the demonstrated role of mitochondrial dysfunction in the disease. It is also notable that genes associated with the RNA function and translational regulation were identified in all three gene ontology classifications. The significant over-representation of ribonucleoprotein complex (within cellular components), of RNA binding (within molecular function), and of ribosome biogenesis (within biological function) supports a role for dysfunction in RNA metabolism and protein translation in addition to the basic mitochondrial respiratory involvement in FRDA.

Table 2
Gene ontology categorization and EASE analysis

Quantitative RT-PCR confirmation of differential expression

Confirmation on KIKO samples

The expression of 25 genes, chosen to represent a cross section of genes expressed at different levels and regions, was tested on RNA extracted from an independent set of mutant and WT animals, by means of qRT-PCR, so as to provide an independent confirmation of the microarray results (Fig. 3). qRT-PCR data confirmed the microarray data for 18/25 (72%) of the genes.

Fig. 3
Microarray and qRT-PCR data in KIKO mice and FRDA fibroblasts. Microarray data were confirmed through real-time quantitative PCR. Samples from at least 3 distinct animals and controls were tested. A subset of genes was tested on three fibroblast cell ...

Confirmation on FRDA fibroblast cell lines

We next tested the expression levels of the human homologs of 11 genes on RNA extracted from three fibroblast cell lines from FRDA patients. 7/11 genes (64%) showed the same changes in human FRDA fibroblasts (Fig. 3). Since fibroblasts are not neural tissue, such a level of confirmation was close to what might be expected a priori, based on experience from our group and others, when confirming changes in different cell types and across different methods (array vs. qRT-PCR).


The goal of this microarray study was to identify a biochemical phenotype secondary to a significant reduction in frataxin levels in clinically relevant brain regions, prior to the onset of any neurodegeneration or clinical phenotype. This avoids confounding factors, as cell loss or reactive changes occurring during overt neurodegeneration. Over 200 differentially expressed genes involved in several pathways were identified. Quantitative RT-PCR was used in independent KIKO mouse samples and in fibroblasts from FRDA patients and confirmed a significant proportion of these changes. A subset of genes showed region-specific changes, mostly involving the cervical spinal cord, which is a region heavily involved in the human disease. Consistent with the subtle biochemical phenotype expected, the magnitude of the changes observed was small, and in many cases likely providing a compensatory mechanism to counteract cellular stress induced by reduced frataxin.

Current pathogenetic theories propose a role of frataxin in ISC assembly (Acquaviva et al., 2005; Muhlenhoff et al., 2002; Stehling et al., 2004), in the activation of stress pathway (Pianese et al., 2002), and in iron metabolism (Cavadini et al., 2002). Microarray studies of Δyfh1 yeast strains (knockout for the yeast frataxin homolog) showed increased expression of genes involved in iron level regulation (Foury and Talibi, 2001). In the first study involving human cells, Tan et al. reported altered expression of several classes of genes, including amino acid metabolism, apoptosis and signal transduction; these authors focused on the involvement of the sulfur amino acid pathway (which is connected to the ISC biosynthetic pathway) and confirmed this finding through functional experiments (Tan et al., 2003). In two recent studies, cardiac and liver tissues from a conditional frataxin knock-out were studied with microarrays, and showed expression changes in genes involved in amino acid (Seznec et al., 2005) and heme metabolism (Schoenfeld et al., 2005). In the present study, GO analysis and literature review showed that the genes identified are involved in nucleic acid and protein metabolism, signal transduction, stress response, and nucleic acid binding. The over-representation of mitochondria-related transcripts within the subset of the differentially expressed genes supports an involvement of mitochondrial pathways secondary to the deficiency of frataxin, a nuclear-encoded mitochondrial protein. Thus, our study adds further evidence supporting the mitochondrial and amino acid metabolism involvement, the activation of stress pathways, and little involvement of iron metabolism-related genes in the early steps of FRDA pathogenesis.

OX-REDOX chemistry and disease pathophysiology

Oxidative stress plays an important role in the pathogenesis of FRDA (Puccio and Koenig, 2002), and this may be linked to the ISC biosynthesis defect. Antioxidant defenses have been reported reduced in FRDA cells (Chantrel-Groussard et al., 2001; Jiralerspong et al., 2001), and increased in transgenic cells overexpressing frataxin (Shoichet et al., 2002). Glutathione reductase catalyzes the NADPH-dependent reduction of oxidized glutathione (GSSG) to glutathione (GSH), and is essential in maintaining adequate levels of reduced GSH. Levels of mRNA coding for glutathione reductase 1 are reduced in the cervical spinal cord of frataxin deficient mice, and in fibroblasts from patients. This observation is intriguing, since a lower activity of this enzyme has been reported in the blood of FRDA patients (Helveston et al., 1996) and higher levels of GSSG have been found in frataxin-deficient cells (Tan et al., 2003). The gene NHL repeat containing 2 has a thioredoxin domain and its transcript is downregulated across all the CNS regions and in FRDA fibroblasts. Thioredoxins play a key role in maintaining proteins in their reduced state, and in defense against oxidative stress (Arner and Holmgren, 2000). Taken together, our data support the role of an early deficiency in the oxidative stress-related pathways in this animal model, and offer a contribution to the general debate about the role of oxidative stress in neurodegeneration (Andersen, 2004), especially after two recent studies respectively supporting (Sturm et al., 2005) and suggesting a revision (Seznec et al., 2005) of the concept of FRDA as a paradigm for neurodegenerative diseases due to oxidative stress.

The involvement of the stress-pathway response

The mitogen-activated protein kinase (MAPK) signaling cascade is implicated in several cellular processes, including regulation of gene expression in response to environmental stress (Chang and Karin, 2001). A hyperactive stress pathway, involving the mitogen activated protein kinase kinase 4 (MAP2K4) and the c-JUN N-terminal kinase was reported in FRDA fibroblasts and in a FRDA foetus, suggesting an early role in the disease pathogenesis (Pianese et al., 2002). Map4k5, coding for a member of the MAPK family, is upregulated in frataxin deficient mice. AVED – a human neurodegenerative disorder caused by mutations in the TTPA gene, coding for the α-tocopherol transfer protein – is often clinically indistinguishable from FRDA (Ben Hamida et al., 1993), suggesting some common pathogenetic pathways. Thus, it is striking that the gene expression profile identified here shares some analogies with that of vitamin E deficiency mice. Map2k3, a member of the MAPK cascade, is upregulated in the liver of an AVED mouse model (Gohil et al., 2003), and members of the same family have been identified as vitamin E sensitive transcripts (Roy et al., 2002). RAR-related orphan receptor alpha (downregulated in brains of KIKO mice and in FRDA fibroblasts) is involved in the lipid metabolism and in protection against age-related degenerative processes (Boukhtouche et al., 2004), and is strongly downregulated in the cortex of Ttpa−/− mice. Of note, the spontaneous staggerer mouse is caused by a mutation in this gene, and is associated with ataxia and cerebellar degeneration (Hamilton et al., 1996).

RNA and protein metabolism

The role of RNA and protein metabolism evident in the GO analysis from KIKO mice is supported by other reports in literature. In addition to cysteinyl-tRNA synthetase (upregulated in this study), 3 other tRNA-synthetases (Gln-, Asn-, and Ala-tRNA synthetase) have been reported as upregulated in hearts of frda mutants (Seznec et al., 2005), and another (seryl-tRNA synthetase) was previously reported as downregulated in FRDA lymphoblasts (Tan et al., 2003), suggesting an involvement of intracellular amino acid metabolism in the pathogenesis of the disease.

Importance of a ‘microarray phenotype’ in absence of overt neurodegeneration

The presence of a gene expression phenotype raises questions about the absence of a clinical phenotype in this model. The transcriptional profile in KIKO mice may be involved in a compensatory response aimed at maintaining cellular function and integrity, or constitute an early step in disease pathogenesis. In the first case, the compensatory changes would underlie the absent phenotype; in the second, the lifespan of frataxin deficient mice may be too short to detect a clinical phenotype. Therefore, the possibility of a very late-onset disease in this model (or an undetectable underlying pathologic process) should be considered. The conditional FRDA mouse model shows signs of neurological impairment about 6 months after the knocking-down of frataxin (Simon et al., 2004), and the knockout mouse for α-TTP (α-TTP−/−) – a model of a late-onset, slowly progressive neuronal degeneration due to chronic oxidative stress – did not present clinical or pathological phenotype until after 1 year of age (Yokota et al., 2001). But, similar to our observations here, a gene expression phenotype could be detected in a similar model at 12–16 weeks of age (Gohil et al., 2003). Thus, the presence of a detectable phenotype is related to many factors, including age, genetic background, and type of phenotypic analysis. Although fibroblasts are not known to be involved in human disease, they provide an accessible source for comparing in human tissue changes found in mouse. Thus, we do not expect all the changes in brain to be measured in fibroblasts, but we were able to confirm a subset of the genes changing in KIKO mice in FRDA fibroblasts. This provides some additional support for the use of mouse models in this disorder, and additional candidate genes for further investigation through functional studies. In addition, the definition of a core of genes changing due to frataxin deficiency can be useful in the evaluation of small molecules with possible therapeutic value: looking for candidate drugs able to revert a gene expression phenotype based on 10–15 genes may be more sensitive than relying entirely on frataxin levels.

Some methodological issues should be addressed. Even after a conservative statistical approach, some of the detected changes in a microarray study can always be due to biological variability. However, in this case, we reduced this effect by performing many replicates (8 per brain region), pooling the samples, and confirming the differentially expressed genes with qRT-PCR on animals distinct from those tested in the microarray study. A concordance of 72% between qRT-PCR and microarray data is well within typical levels of confirmation, especially using independent samples. In addition, nearly 10% of the observed changes were confirmed in this manner by qRT-PCR, a large cross section of the data. It should be emphasized that the small magnitude of the detected changes challenges the sensitivity of both microarray and qRT-PCR techniques. However, the rate of concordance between the two techniques was very reasonable, supporting the validity of the results.

In conclusion, the identification of a gene expression profile associated with reduced frataxin levels in this animal model provides valuable insights for further studies aimed at both understanding the earliest molecular events in FRDA pathogenesis, and in setting up in vitro tools to evaluate new therapeutic strategies.


We thank Coriell Cell Repositories for providing FRDA and control fibroblast cell lines, and Arnulf Koeppen, MD for critically reading the manuscript.

This work was supported by a research grant from Friedreich’s Ataxia Research Alliance/MDA Seek-A-Miracle to GC and DHG, the William Smith Memorial fund gift to DHG, and by the National Institutes of Health (grant no. NS34192) to MP.


  • Acquaviva F, De Biase I, Nezi L, Ruggiero G, Tatangelo F, Pisano C, et al. Extra-mitochondrial localisation of frataxin and its association with IscU1 during enterocyte-like differentiation of the human colon adenocarcinoma cell line Caco-2. J Cell Sci. 2005;118:3917–3924. [PubMed]
  • Andersen JK. Oxidative stress in neurodegeneration: cause or consequence? Nat Med. 2004;10(Suppl):S18–S25. [PubMed]
  • Arner ES, Holmgren A. Physiological functions of thioredoxin and thioredoxin reductase. Eur J Biochem. 2000;267:6102–6109. [PubMed]
  • Ben Hamida C, Doerflinger N, Belal S, Linder C, Reutenauer L, Dib C, et al. Localization of Friedreich ataxia phenotype with selective vitamin E deficiency to chromosome 8q by homozygosity mapping. Nat Genet. 1993;5:195–200. [PubMed]
  • Bidichandani SI, Ashizawa T, Patel PI. The GAA triplet-repeat expansion in Friedreich ataxia interferes with transcription and may be associated with an unusual DNA structure. Am J Hum Genet. 1998;62:111–121. [PubMed]
  • Boukhtouche F, Mariani J, Tedgui A. The “CholesteROR” protective pathway in the vascular system. Arterioscler Thromb Vasc Biol. 2004;24:637–643. [PubMed]
  • Campuzano V, Montermini L, Molto MD, Pianese L, Cossee M, Cavalcanti F, et al. Friedreich’s ataxia: autosomal recessive disease caused by an intronic GAA triplet repeat expansion. Science. 1996;271:1423–1427. [PubMed]
  • Campuzano V, Montermini L, Lutz Y, Cova L, Hindelang C, Jiralerspong S, et al. Frataxin is reduced in Friedreich ataxia patients and is associated with mitochondrial membranes. Hum Mol Genet. 1997;6:1771–1780. [PubMed]
  • Cavadini P, O’Neill HA, Benada O, Isaya G. Assembly and iron-binding properties of human frataxin, the protein deficient in Friedreich ataxia. Hum Mol Genet. 2002;11:217–227. [PubMed]
  • Chang L, Karin M. Mammalian MAP kinase signalling cascades. Nature. 2001;410:37–40. [PubMed]
  • Chantrel-Groussard K, Geromel V, Puccio H, Koenig M, Munnich A, Rotig A, et al. Disabled early recruitment of antioxidant defenses in Friedreich’s ataxia. Hum Mol Genet. 2001;10:2061–2067. [PubMed]
  • Cossee M, Puccio H, Gansmuller A, Koutnikova H, Dierich A, LeMeur M, et al. Inactivation of the Friedreich ataxia mouse gene leads to early embryonic lethality without iron accumulation. Hum Mol Genet. 2000;9:1219–1226. [PubMed]
  • Foury F, Talibi D. Mitochondrial control of iron homeostasis. A genome wide analysis of gene expression in a yeast frataxin-deficient strain. J Biol Chem. 2001;276:7762–7768. [PubMed]
  • Gerber J, Muhlenhoff U, Lill R. An interaction between frataxin and Isu1/Nfs1 that is crucial for Fe/S cluster synthesis on Isu1. EMBO Rep. 2003;4:906–911. [PubMed]
  • Geschwind DH. Mice, microarrays, and the genetic diversity of the brain. Proc Natl Acad Sci U S A. 2000;97:10676–10678. [PubMed]
  • Gohil K, Schock BC, Chakraborty AA, Terasawa Y, Raber J, Farese RV, Jr, et al. Gene expression profile of oxidant stress and neurodegeneration in transgenic mice deficient in alpha-tocopherol transfer protein. Free Radical Biol Med. 2003;35:1343–1454. [PubMed]
  • Hamilton BA, Frankel WN, Kerrebrock AW, Hawkins TL, FitzHugh W, Kusumi K, et al. Disruption of the nuclear hormone receptor RORalpha in staggerer mice. Nature. 1996;379:736–739. [PubMed]
  • Helveston W, Hurd R, Uthman B, Wilder BJ. Abnormalities of glutathione peroxidase and glutathione reductase in four patients with Friedreich’s disease. Mov Disord. 1996;11:106–107. [PubMed]
  • Hosack DA, Dennis G, Jr, Sherman BT, Lane HC, Lempicki RA. Identifying biological themes within lists of genes with EASE. Genome Biol. 2003;4:R70. [PMC free article] [PubMed]
  • Jiralerspong S, Ge B, Hudson TJ, Pandolfo M. Manganese superoxide dismutase induction by iron is impaired in Friedreich ataxia cells. FEBS Lett. 2001;509:101–105. [PubMed]
  • Karsten SL, Geschwind DH. Gene expression analysis using cDNA microarrays. Curr Prot Neurosci. 2002:4.28.1–4.28.28.
  • Karsten SL, Van Deerlin VM, Sabatti C, Gill LH, Geschwind DH. An evaluation of tyramide signal amplification and archived fixed and frozen tissue in microarray gene expression analysis. Nucleic Acids Res. 2002;30:E4. [PMC free article] [PubMed]
  • Lamarche JB, Lemieux B, Lieu HB. The neuropathology of “typical” Friedreich’s ataxia in Quebec. Can J Neurol Sci. 1984;11:592–600. [PubMed]
  • Liang M, Briggs AG, Rute E, Greene AS, Cowley AW., Jr Quantitative assessment of the importance of dye switching and biological replication in cDNA microarray studies. Physiol Genomics. 2003;14:199–207. [PubMed]
  • Lutz T, Westermann B, Neupert W, Herrmann JM. The mitochondrial proteins Ssq1 and Jac1 are required for the assembly of iron sulfur clusters in mitochondria. J Mol Biol. 2001;307:815–825. [PubMed]
  • Miranda CJ, Santos MM, Ohshima K, Smith J, Li L, Bunting M, et al. Frataxin knockin mouse. FEBS Lett. 2002;512:291–297. [PubMed]
  • Muhlenhoff U, Richhardt N, Ristow M, Kispal G, Lill R. The yeast frataxin homolog Yfh1p plays a specific role in the maturation of cellular Fe/S proteins. Hum Mol Genet. 2002;11:2025–2036. [PubMed]
  • Pfaffl MW, Horgan GW, Dempfle L. Relative expression software tool (REST) for group-wise comparison and statistical analysis of relative expression results in real-time PCR. Nucleic Acids Res. 2002;30:e36. [PMC free article] [PubMed]
  • Pianese L, Busino L, De Biase I, De Cristofaro T, Lo Casale MS, Giuliano P, et al. Up-regulation of c-Jun N-terminal kinase pathway in Friedreich’s ataxia cells. Hum Mol Genet. 2002;11:2989–2996. [PubMed]
  • Pianese L, Turano M, Lo Casale MS, De Biase I, Giacchetti M, Monticelli A, et al. Real time PCR quantification of frataxin mRNA in the peripheral blood leucocytes of Friedreich ataxia patients and carriers. J Neurol Neurosurg Psychiatry. 2004;75:1061–1063. [PMC free article] [PubMed]
  • Puccio H, Koenig M. Friedreich ataxia: a paradigm for mitochondrial diseases. Curr Opin Genet Dev. 2002;12:272–277. [PubMed]
  • Puccio H, Simon D, Cossee M, Criqui-Filipe P, Tiziano F, Melki J, et al. Mouse models for Friedreich ataxia exhibit cardiomyopathy, sensory nerve defect and Fe–S enzyme deficiency followed by intramitochondrial iron deposits. Nat Genet. 2001;27:181–186. [PubMed]
  • Ristow M, Mulder H, Pomplun D, Schulz TJ, Muller-Schmehl K, Krause A, et al. Frataxin deficiency in pancreatic islets causes diabetes due to loss of beta cell mass. J Clin Invest. 2003;112:527–534. [PMC free article] [PubMed]
  • Roy S, Lado BH, Khanna S, Sen CK. Vitamin E sensitive genes in the developing rat fetal brain: a high-density oligonucleotide microarray analysis. FEBS Lett. 2002;530:17–23. [PMC free article] [PubMed]
  • Santos MM, Miranda CJ, Levy JE, Montross LK, Cossee M, Sequeiros J, et al. Iron metabolism in mice with partial frataxin deficiency. Cerebellum. 2003;2:146–153. [PubMed]
  • Schoenfeld RA, Napoli E, Wong A, Zhan S, Morin D, Buckpitt AR, et al. Frataxin deficiency alters heme pathway transcripts and decreases mitochondrial heme metabolites in mammalian cells. Hum Mol Genet. 2005;14:3787–3799. [PubMed]
  • Serra HG, Byam CE, Lande JD, Tousey SK, Zoghbi HY, Orr HT. Gene profiling links SCA1 pathophysiology to glutamate signaling in Purkinje cells of transgenic mice. Hum Mol Genet. 2004;13:2535–2543. [PubMed]
  • Seznec H, Simon D, Bouton C, Reutenauer L, Hertzog A, Golik P, et al. Friedreich ataxia, the oxidative stress paradox. Hum Mol Genet. 2005;14:463–474. [PubMed]
  • Shoichet SA, Baumer AT, Stamenkovic D, Sauer H, Pfeiffer AF, Kahn CR, et al. Frataxin promotes antioxidant defense in a thiol-dependent manner resulting in diminished malignant transformation in vitro. Hum Mol Genet. 2002;11:815–821. [PubMed]
  • Simon D, Seznec H, Gansmuller A, Carelle N, Weber P, Metzger D, et al. Friedreich ataxia mouse models with progressive cerebellar and sensory ataxia reveal autophagic neurodegeneration in dorsal root ganglia. J Neurosci. 2004;24:1987–1995. [PubMed]
  • Sipione S, Rigamonti D, Valenza M, Zuccato C, Conti L, Pritchard J, et al. Early transcriptional profiles in huntingtin-inducible striatal cells by microarray analyses. Hum Mol Genet. 2002;11:1953–1965. [PubMed]
  • Stehling O, Elsasser HP, Bruckel B, Muhlenhoff U, Lill R. Iron–sulfur protein maturation in human cells: evidence for a function of frataxin. Hum Mol Genet. 2004;13:3007–3015. [PubMed]
  • Sturm B, Bistrich U, Schranzhofer M, Sarsero JP, Rauen U, Scheiber-Mojdehkar B, et al. Friedreich’s ataxia: no changes in mitochondrial labile iron in human lymphoblasts and fibroblasts: a decrease in antioxidative capacity? J Biol Chem. 2005;280:6701–6708. [PubMed]
  • Tan G, Napoli E, Taroni F, Cortopassi G. Decreased expression of genes involved in sulfur amino acid metabolism in frataxin-deficient cells. Hum Mol Genet. 2003;12:1699–1711. [PubMed]
  • Watts JA, Morley M, Burdick JT, Fiori JL, Ewens WJ, Spielman RS, et al. Gene expression phenotype in heterozygous carriers of ataxia telangiectasia. Am J Hum Genet. 2002;71:791–800. [PubMed]
  • Yang YH, Dudoit S, Luu P, Lin DM, Peng V, Ngai J, et al. Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation. Nucleic Acids Res. 2002;30:e15. [PMC free article] [PubMed]
  • Yokota T, Igarashi K, Uchihara T, Jishage K, Tomita H, Inaba A, et al. Delayed-onset ataxia in mice lacking alpha -tocopherol transfer protein: model for neuronal degeneration caused by chronic oxidative stress. Proc Natl Acad Sci U S A. 2001;98:15185–15190. [PubMed]
  • Yoshihara T, Ishigaki S, Yamamoto M, Liang Y, Niwa J, Takeuchi H, et al. Differential expression of inflammation- and apoptosis-related genes in spinal cords of a mutant SOD1 transgenic mouse model of familial amyotrophic lateral sclerosis. J Neurochem. 2002;80:158–167. [PubMed]