Idebenone: A three-level demonstration of efficacy Following the demonstration in 1999 of the efficacy of idebenone in an a-cellular system prepared from human heart tissue, several partners of the consortium have concur to study its efficacy using frataxin-deficient (or lacking) intact cultured cells, mice, and human samples. a) Intact cultured cells In the course of the project, partners 3 and 5 have developed and used complementary oxidative stress systems allowing to differentially trigger cell death in control and frataxin-deficient human skin fibroblasts. Although the mechanism of the hypersensitivity of these latter cells to oxidative stress remains to be firmly established, it appears that on one hand these oxidative stresses permitted to readily test for protection potentially afforded by any compound of interest, and on the other hand mimicked part of the injury observed in the course of the disease. Using these models, it was established that idebenone fully protected frataxin-deficient cells from death induced by the used oxidative stresses. b) Mouse models A mouse model for Friedreich ataxia was generated by partner 1 using mice obtained by crossing homozygous animals for a conditional allele of frataxin with animal heterozygous for a deletion of exon 4 of frataxin gene and carrying a Cre transgene under the control of the muscle creatine kinase. These mice lack frataxin in their striated muscle and develop from 4 to 8 wks a spectacular and progressive hypertrophic cardiomyopathy with fatal outcome. Enzyme analysis carried out in the heart tissue of these mice showed a loss of iron-sulphur cluster-containing protein (ISP) activity, as specifically observed in Friedreich ataxia patients. These animals were given idebenone (90 mg/kg/d from 3 to 8 weeks). As the result of this treatment, a simultaneous slow down of cardiomyopathy installation and increased life expectancy were noticed. However, it was found that the improvement of these parameters was not directly related to protection of ISP, since iron-sulphur cluster containing respiratory chain complex II was as severely impaired in both treated and untreated animals. This study established that idebenone was actually able to protect heart function even under condition where frataxin is fully absent. Lower doses (30 mg/kg/d and 10 mg/kg/d) were barely or not active, respectively. c) Human samples Following the open trial performed with idebenone in France in a large cohort of patients (less than 60), partner 3 had the opportunity to compare ISP activity in endomyocardial biopsies from one patient before and after 5 years idebenone treatment (5 mg/kg/d). It was observed an amazing restoration of ISP activities which values were normalized in the heart following the treatment. This indicated that residual frataxin, always present in patient tissue, is sufficient to restore ISP synthesis when idebenone was given to patient, providing a simple explanation for the positive effect of idebenone on the course of cardiomyopathy in Friedreich ataxia patients. Altogether these data indicate that (1) idebenone is a powerful antioxidant able to counteract various oxidative insults in situ in living cells, as initially suggested in the a-cellular system, (2) presumably by doing so, it counteracts heart hypertrophy in mouse model and patients, albeit with potentially different mechanism, and (3), if residual frataxin is sufficient, it simultaneously allows the restoration of ISP protein activity. References. Geromel V, Darin N, Chretien D, Benit P, DeLonlay P, Rotig A, Munnich A, Rustin P. Coenzyme Q(10) and idebenone in the therapy of respiratory chain diseases: rationale and comparative benefits. Mol Genet Metab. 2002, Hausse AO, Aggoun Y, Bonnet D, Sidi D, Munnich A, Rotig A, Rustin P. Idebenone and reduced cardiac hypertrophy in Friedreich's ataxia. Heart. 2002 Apr;87(4):346-9. Jauslin ML, Wirth T, Meier T, Schoumacher F. A cellular model for Friedreich Ataxia reveals small-molecule glutathione peroxidase mimetics as novel treatment strategy. Hum Mol Genet. 2002;11:3055-63 Puccio H., Simon D., Monassier L., Seznec H., Criqui-Filipe P., Coumaros G., Gansmuller A., Rustin P., Koenig M. (2004) Dose-dependent protective effects of idebenone against the cardiac phenotype in mouse model for Friedreich ataxia. Rotig A, Sidi D, Munnich A, Rustin P. Molecular insights into Friedreich's ataxia and antioxidant-based therapies. Trends Mol Med. 2002 May;8(5):221-4. Rustin P, Bonnet D, Rötig A, Munnich A, Sidi D. (2004) Idebenone restores mitochondrial respatory chain enzyme activities in the cardiac muscle in Friedreich�s ataxia. Neurology (in press) Rustin P, Rotig A, Munnich A, Sidi D. Heart hypertrophy and function are improved by idebenone in Friedreich's ataxia. Free Radic Res. 2002 Apr;36(4):467-9. Rustin P. The use of antioxidants in Friedreich's ataxia treatment.Expert Opin Investig Drugs. 2003 Apr;12(4):569-75.
At the beginning of this project, the function of frataxin was unknown. Yeast proved a very useful model organism for searching the function of frataxin. Thus the fundamental results we have obtained during the course of this project can be applied to humans. Deletion of the YFH1 yeast gene is associated with loss of respiration, excess mitochondrial iron accumulation and defects in iron-sulphur proteins. In addition, cells exhibit increased sensitivity to oxidative reagents. It was first believed that the defects were caused by oxidative damage resulting from free radicals generated by excess iron. However, we were able to prevent mitochondrial iron overload by growing yeast cells in low iron media without restoring the activity of iron-sulphur cluster proteins. These data suggested that the primary cause of the mitochondrial dysfunction was a specific defect in the activity of iron-sulphur cluster proteins resulting from impaired iron-sulphur cluster biosynthesis. This hypothesis was consistent with the discovery by Lill and co-workers that iron-sulphur clusters are synthesized in the mitochondria. In this work we have established that frataxin is directly involved in the biosynthesis of iron-sulphur clusters. 1) Preventing mitochondrial iron overload by genetical and physiological means does not restore iron-sulphur protein activity. a)When yfh1 deleted cells grow on glycerol mitochondria respire, do not accumulate iron but have reduced activity of aconitase, a 4Fe-4S protein. Thus even in absence of any obvious oxidative damage Fe-S cluster proteins are defective. b) We introduced mrs3/mrs4 deleted alleles in a yfh1 deleted strain. MRS3/MRS4 encode mitochondrial carriers playing an important role in mitochondrial iron uptake. No mitochondrial iron accumulation is observed in the triple yfh1 mrs3 mrs4 deleted strain, yet aconitase activity is still lower than in the yfh1-deleted strain. However, in a wild-type strain, deletion of MRS3/MRS4 has only a slight effect on aconitase. We concluded that iron-sulphur protein deficiency is a specific defect associated with loss of frataxin. Mitochondrial iron overload and oxidant sensitivity could be a secondary effect. This work has been disseminated by two publications (cited 30 and 10 times respectively) Foury, F., and Talibi, D. (2001) J Biol Chem 276, 7762–7768Foury, F., and Roganti, T. (2002) J Biol Chem 277, 24475-24483 2) Iron-sulphur cluster synthesis is defective in organelle. We have isolated mitochondria from wild type and yfh1 deleted strains and followed the kinetics of incorporation of Fe-S cluster into Yah1p, a mitochondrial 2Fe-2S protein, imported into mitochondria in the presence of cysteine. The apo- and holo-Yah1p are distinguished by their mobility in native acrylamide gel electrophoresis. We found that the incorporation rate of the iron-sulphur cluster into Yah1p is substantially decreased in the yfh1-deleted strain. This work performed by Dr. G. Duby, a post-doctoral fellow, has been disseminated by one publication. Duby, G., Foury, F., Ramazzotti, A., Herrmann, J., and Lutz, T. (2002) Hum. Mol. Genet. 11, 2635-2643. 3) Frataxin is a functional and direct partner of ISU, the essential scaffold protein for iron-sulphur cluster assembly. We have searched for partners of frataxin, using a blind genetic search based on a synthetic lethal screen on iron media. We firstly isolated mild yfh1 mutants killed by iron at 37°C. To identify genes functionally related to frataxin, we performed a synthetic lethal screen for extragenic mutations that are lethal with the yfh1 G107D mutation on iron at 28°C. We performed an in vivo chemical mutagenesis and identified a double mutation in isu1 (G60D-M141V) as responsible for the synthetically lethal trait. Isu1p, a central protein in the iron-sulphur cluster machinery, plays the role of a scaffold to assemble a 2Fe-2S cluster from iron and sulphide. Using cross-linking in intact mitochondria we showed cross-linking between yeast frataxin and Isu1p, strongly suggesting that these two proteins are physically associated to perform iron-sulphur cluster synthesis. These data show that frataxin might be an iron chaperone mediating iron delivery to Isu1p, and thus plays a direct role in the iron-sulphur cluster biosynthesis. This work by Ramazzotti, a senior scientist appointed for the whole duration of this contract has been accepted for publication. Ramazzotti A. Vanmansart, V. and Foury, F. (2004) FEBS Letters conclusion The work accomplished during this EC contract has brought essential fundamental knowledge, relative to the function of frataxin in yeast that is applicable to humans, and thus can be considered as an important basis for medical and therapeutical purposes.
Sense and anti-sense frataxin adenoviral vectors as tools to study frataxin function in mammalian cells
Studies of the neuronal cell death mechanisms in FRDA depend on the availability of cellular and animal models. Following the hypothesis generated in yeast that a depletion of FRDA may increase oxidative stress we studied several in vivo markers of oxidative stress in FRDA patients. 1) Of these, urinary 8-hydroxy-2’-deoxyguanosine (8OH2’dG) concentrations were significantly increased in FRDA patients. Furthermore, treatment with the antioxidant idebenone decreased the concentration of 8OH2’dG in these patients. Therefore, 8OH2’dG is a surrogate biomarker to detect oxidative stress in FRDA patients and may serve to monitor the efficacy of drug treatments 2) The importance of these results was underlined in an editorial. 3) Idebenone has been shown to protect against cardiomyopathy in mice and man. However, it does not have a symptomatic effect on neurological symptoms and is unlikely to have an influence on the rate of disease progression (Alexandra Dürr, Paris, personal communication). To date there is no model for Friedreich’s ataxia to investigate the effects of putative protective drugs in primary postmitotic neurons in culture. Therefore, drugs cannot be screened for their potential to protect against neuronal death in Friedreich’s ataxia. In order to create a cellular model of frataxin deficiency in postmitotic neurons we produced an adenovirus coding for antisense frataxin (AdV-as-frataxin). Transfection of HEK 293 cells, SH-SY5Y cells as well as primary cerebellar granule neurons (CGN) with AdV-as-frataxin led to reduced expression of endogenously or ectopically expressed frataxin mRNA and protein. Transfection with AdV-as-frataxin leads to death of proliferating SH-SY5Y cells without applying any other stressor. This is in line with the failed attempts of the other partners of the consortium to create cellular models in proliferating cells (e.g. fibroblasts) with a complete knock out of frataxin. We next investigated the effects of modulating frataxin expression induced by transfecting postmitotic differentiated cerebellar granule neurons with AdV-as-frataxin or AdV-sense-frataxin. During the first 7 days after transfection with either vector we did not detect a difference in cell viability compared with cells treated with a control vector. Therefore, we exposed cerebellar granule neurons with a reduced frataxin expression to an apoptotic (potassium withdrawal) or excitotoxic (glutamate) stimulus. Virus mediated forced expression of frataxin resulted in enhanced resistance and virus-antisense mediated reduced expression of frataxin in an increased sensitivity of cerebellar granule neurons against potassium withdrawal and glutamate exposure. Fifty percent of the latter neurons were rescued by treatment with idebenone 4) To further characterize the therapeutic potential of AdV-sense-frataxin in the replacement of frataxin we tested the metabolism of ectopically expressed frataxin. Cells were transfected in vitro, the protein was expressed and processed to its active form, presumably by a mitochondrial peptidase. In vivo, after stereotaxic injection of this vector into the brain, we confirmed the expression of frataxin in the striatum of mice and rats by immunohistochemistry. In parallel to the work funded by this grant we developed a model of aging in cerebellar granule neurons that allows us to culture these postmitotic neurons for more than 60 days. In cerebellar granule neurons tranfected with AdV-as-frataxin at day-in-vitro 1 that were then aged for 30 days in culture we observed a 70% reduction of viable neurons compared with aged neurons that had been transfected with a control vector. In summary, we have for the first time created a model to investigate the consequences of frataxin deficiency in differentiated neurons and developed a tool for an efficacious replacement therapy in Friedreich’s ataxia. literature 1. Schulz JB, Dehmer T, Schöls L, Mende H, Hardt C, Vorgerd M, Burk K, Matson W, Dichgans J, Beal MF, Bogdanov MB. Oxidative stress in patients with friedreich ataxia. Neurology 2000;55:1719-1721 2. Schulz JB, Lindenau J, Seyfried J, Dichgans J. Glutathione, oxidative stress and neurodegeneration. Eur. J. Biochem. 2000, 267: 4904-49113. Sherer T, Greenamyre JT. A therapeutic target and biomarker in Friedreich's ataxia. Neurology 2000;55:1600-1601 4. Wick A, Gerhardt E, Gleichmann M, Schulz JB. Consequences of reduced frataxin expression in differentiated and aged neurons. In preparation
During the total period of this program we have established a disease-relevant cellular assay for Friedreich Ataxia, which is based on the inhibition of de novo synthesis of glutathione (GSH). The specific cell death of FRDA derived fibroblast could be used to determine the potency of various agents that were able to prevent cell death. As a standard we referred to the known antioxidant idebenone, currently used for the treatment of the disease related hypertrophic cardiomyopathy.We have identified several classes of potentially bioactive molecules, which could prevent the FRDA-cell death caused by GSH.These include known antioxidant such as idebenone and vitamine E, but also novel molecules such as selenium and small molecule glutathione peroxidase mimetics (1). Since selenium was active in this assay, we next analysed the question whether a misregulation of Se-metabolism might be present in FRDA patients. However, by determining Se-dependent parameters in plasma from FRDA patients such as selenoprotein P and glutathione peroxidase, we found no support for this theory (2). Further we focussed on the question, how the activity/potency of existing antioxidants could be increased. As a result, we found that a combination of certain antioxidants has a stronger effect on the viability of BSO-challenged FRDA fibroblasts than the administration of individual antioxidant. In particular, we found that the concomitant application of vitamin E and idebenone is approximately for times more potent than either one alone. This synergistic pattern is seen both with equimolar mixtures of idebenone/vitamin E, but also with novel synthetic molecules consisting of the active moieties from both antioxidants (3). As an alternative approach, we compared idebenone to a mitochondrial-targeted version of idebenone (MitoQ) and found, that this was several hundred times more potent that the non targeted version. This is of relevance, since to date idebenone shows relatively low efficacy in FRDA patients (4). In summary, we demonstrate several options how FRDA treatment could be improved; based both on existing therapy and on novel pharmacophores. Our findings have been published in several high-ranking journals, gave rise to one PhD thesis and were filed for patent. (1) Hum Mol Genet. 2002 Nov 15;11(24):3055-63; (2) Jauslin M. (2003) PhD Thesis, University of Basel; (3) Jauslin M. et al. (manuscript in prep.) (4) Jauslin M. et al. (2003) FASEB J. 2003 Oct;17(13):1972-4.
Prior to the beginning of this project, there was no effective treatment for FRDA that would slow down the progression of the disease. Clinical trials for rare diseases pose the ethical problems of double blind studies where half the patients receives a placebo, and the extreme difficulty to evaluate patients. This is even exacerbated for Friedreich ataxia, because of the slow progression and variable presentation of the disease related to the variable size of the causal expansion mutation. Therefore, the only possibility to confirm effective drugs rapidly is through the use of animal models of the disease. For patients and families, who strongly support this enterprise, and scientists, there was clearly no other alternative. The construction of mice models for FRDA turned out to be particularly difficult, since the human disease is the consequence of very reduced frataxin protein levels, its total absence causing very early embryonic lethality. In addition, the human mutation, a trinucleotide expansion, would cause a disease with onset past the normal life expectancy of the mouse, and all attempts to introduce the FRDA expansion in the mouse have failed to recreate the very large expansions that might be associated with an earlier onset. We chose to recreate a partial loss of function by the conditional knockout strategy, whereby the inactivation of the frataxin gene occurs later during embryonic development or animal life and in restricted tissues. The ensuing frataxin decay then creates a time window during which very reduced frataxin levels and reduced products (Fe-S clusters) perfectly mimick the human pathological situation. The conditional knock-outs appeared to be very good and versatile models that recreate important features of the human disease, including: i) progressive hypertrophic cardiomyopathy without skeletal muscle involvement, ii) progressive ataxia and loss of proprioception without peripheral motor neuropathy, iii) multiple Fe-S dependent enzyme deficiency in affected tissues, iv) time-dependent intramitochondrial iron accumulation. The combination of features was dependent on the use of transgenic recombinase that induces the conditional knockout. We have tested four distinct transgenic recombinases (a ''striated muscle'' recombinase, two ''neuronal'' recombinases and a recombinase expressed both in neurons and heart). Additional combinations can be tested in the future. For example, collaboration with a partner outside from the project resulted in the construction of a model with the specific diabetic aspect of FRDA, by using a specific pancreatic beta cell transgene (Ristow et al J Clin Investigation 2003). Of particular importance, the two ''neuronal'' recombinases that we used were associated with an innovative inducible system, which gives us the additional versatility to modulate the time of frataxin gene deletion at will, and hence also the severity of the disease. This is particularly important for future therapeutic trials, for which both rapid trials on severe models and trials on slowly progressive models more akin to the human disease will be desirable. Our conditional models are the only available mouse models for Friedreich ataxia, despite numerous attempts elsewhere worldwide. The "striated muscle" and the "neuron/heart" models were published in a prestigious journal (Puccio et al. Nature Genetics 2001). They were distributed to a number of laboratories worldwide. The two inducible neuronal models are now in press in the Journal of Neuroscience (Simon et al.) and will be equally distributed, upon request. We used our models to unravel the pathophysiology of the disease. In particular, the neuronal models showed that degeneration of the dorsal root ganglia neurons, the key feature of the human disease, is due to an autophagic process with lipofuscin deposits and does not involve apoptosis or necrosis. This has implications for the design of novel therapeutic avenues. This also strengthened the current therapeutic approach, based on lipophylic antioxidants, such as idebenone, since lipofuscin deposits are known to arise from lipid peroxidation by oxidative stress (as in vitamin E deficiency, also causing spinocerebellar and sensory ataxia). We used the "striated muscle" model to confirm the efficacy of idebenone on the cardiac symptoms of the disease (Seznec et al. submitted), which should translate in the initiation of phase II double blind clinical trials for cardiomyopathy in FRDA in the near future. We plan to extend the study of the effects of idebenone to the inducible neuronal models. We also initiated the use of the "striated muscle" model to assess the effect of iron content in the food and the effect of novel antioxidants, such as MnTBAP, a superoxide dismutase mimetics, and MitoQ, a mitophylic idebenone derivative. Finally, all our models are available for novel therapeutic approaches, including gene therapy and protein replacement therapy.