Introduction

Similar to many other neurodegenerative diseases, Friedreich's ataxia (FA), the most common of the hereditary ataxias, remains incurable and treatment options are limited. Thus there is a pressing need to explore novel therapeutic strategies such as gene therapy. The discovery that mutations in FRDA, the frataxin gene, cause FA¹ has played a key role in advancing knowledge of the molecular mechanism of the disease and opens the way for developing gene-based treatments. Human frataxin is a 210–amino acid protein, although an alternative splice variant has been identified encoding a 196–amino acid form, which differs from the main isoform from amino acid 160 onward.² In FA, the most common pathological mutation consists of a GAA triplet expansion within intron 1 of FRDA¹ that reduces messenger RNA expression to extremely low or undetectable levels in homozygotes. A few patients are heterozygotes for GAA expansions but also have point mutations in the other FRDA allele.^3,4,5,6 It is now generally accepted that FA is due to loss of frataxin function,⁷ and as the introduction of frataxin transgenes in normal mice has no adverse effects,^8,9,10 this disorder is a good model for gene rescue therapy.

The typical neuropathology of FA¹¹ affects the dorsal root ganglia, spinal cord, medulla oblongata, and the dentate nucleus of the cerebellum, with significant atrophy of the superior cerebellar peduncles. Advanced cases are also characterized by atrophy of Purkinje cells in the cerebellar cortex. For gene delivery into the nervous system, a number of viral vectors, such as those derived from herpes simplex virus type 1 (HSV-1), lentivirus, and adeno-associated virus, show great potential.¹² A previous study of FA gene therapy reported the construction of lentivirus and adeno-associated virus vectors encoding human frataxin complementary DNA (cDNA) and their ability partially to correct the sensitivity of FA patient fibroblasts to oxidative stress,¹³ but the use of these vectors in neuronal tissue was not addressed. Although gene transfer experiments with lentiviral and adeno-associated viral vectors are yielding encouraging results in other neuronal systems, HSV-1 vectors possess several advantageous features: they are non-integrating vectors, they are able to accommodate large amounts of DNA (up to 150 kilobases), and because HSV-1 is neurotropic, they have high infectivity in a wide range of neuronal types. We have previously used HSV-1 amplicon vectors for genetic manipulations in the nervous system^14,15 and, in particular, for gene transfer into the rat inferior olive,¹⁶ a nucleus in the medulla with a key role in motor coordination.¹⁷ To address the neurological aspects of FA more rapidly, in this study we have used HSV-1 amplicon vectors to deliver CRE recombinase into the same brain area of conditional (floxed) FRDA transgenic mice (hereafter referred to as loxP[frda] mice),¹⁸ thus provoking by frda gene excision an anatomically localized neuronal lesion with the known phenotypic effect of motor coordination deficit,¹⁷ which can be quantified easily by tests such as the rotarod assay.¹⁹ Using this model we have gone on to perform gene rescue with an HSV-1 amplicon vector encoding human frataxin cDNA and show that neuronal function can be recovered by frda gene therapy at least 1 month after lesion by frataxin depletion.

Results

Generation of loxP[frda] mice and HSV amplicon viral vectors expressing CRE and frataxin

To create the tools for manipulating frataxin levels in vivowe first bred mice homozygous for a conditional frataxin allele¹⁸ (Frda^L3/L3), hereafter referred to as loxP[frda] mice, as described in the Materials and Methods; second, we generated the HSV-1 amplicon vectors (Figure 1) pHLC and pHF, which respectively encode cDNAs expressing CRE recombinase and human frataxin tagged with the FLAG epitope²⁰ (Figure 1a). As a control vector we used pHSVlac,²¹ an amplicon vector that expresses Escherichia coli -galactosidase, which is also expressed by pHLC. The lacZ and frataxin transgenes were expressed from the HSV-1 IE4/5 promoter, and CRE recombinase was expressed from the cytomegalovirus immediate early promoter. Transgene expression was confirmed by transient transfection of the plasmid constructs in Vero cells (Figure 1b) and infection by the packaged HSV-1 amplicon vectors in vivo in loxP[frda] mice (Figure 1d and e). We have previously developed an in vivo strategy to deliver HSV-1 amplicon vectors to brainstem by stereotaxic injection into the adult rat inferior olive.¹⁶ To extend this model system to studies in transgenic mice, we first characterized the spread of injected solution in the mouse brain using the neuronal tracer Fluoro-ruby (Figures 1c). We next examined the expression of pHLC vector injected in vivo in the medulla of loxP[frda] mice. At low vector dosage (3 l of a stock with a titer of 5 10⁶ infectious vector units/ml) transgene expression was hardly detectable in tissue sections but -galactosidase expression was clearly detected (Figure 1d) using higher vector doses (3 l of a stock with a titer of 6.7 10⁷ infectious vector units/ml) and the vector tissue distribution was similar to that marked by Fluoro-ruby. Similarly, transgene expression from the pHF vector was difficult to detect using unconcentrated stocks (3 l of a stock with a titer of 4.8 10⁶ infectious vector units/ml) but clearly discernible by FLAG immunohistochemistry of brainstem sections (Figure 1e) injected using higher doses (3 l of a stock with a titer of 8.7 10⁷ infectious vector units/ml).

Figure 1.

Characterization of herpes simplex virus type 1 (HSV-1) amplicon vectors. (a) Schematic representation of the vectors expressing -galactosidase (pHSVlac: lacZ = blue arrow), -galactosidase and CRE recombinase (pHLC: lacZ = blue arrow, CRE = green arrow), and human frataxin tagged with the FLAG epitope at the C-terminus (pHF: frda = red arrow, FLAG = purple rectangle). The Escherichia colilacZ and human frda complementary DNA (cDNA) transgenes were expressed from the HSV1 IE4/5 promoter, and CRE recombinase was expressed from the cytomegalovirus immediate early promoter. In addition to elements for propagation in bacteria (not shown), each amplicon plasmid contained an HSV-1 origin of replication oriS (S), and the HSV-1 packaging signal "a." (b) Transgene expression observed following transfection of the plasmids in Vero cells. Expression of -galactosidase from the plasmids pHSVlac and pHLC was detected using the substrate X-gal, and expression of the frataxin transgene from pHF was detected by FLAG immunostaining using peroxidase-conjugated secondary antibody. Control untransfected Vero cells were negative for FLAG staining. (c) Injection of Fluoro-ruby (red) into the brainstem in the medulla oblongata. Sections from a mouse sacrificed 1 day after injection were stained by indirect immunofluorescence using an anti-calbindin antibody (green). Scale bar represents 500 m. The approximate location of the field shown in c is indicated by the boxed area. (d) Immunostaining by indirect immunofluorescence for -galactosidase in brainstem neurons 3 days after injection with pHLC (2 10⁵ infectious vector units (ivu)). Scale bar represents 200 m (e) FLAG immunostaining using peroxidase-conjugated secondary antibody in brainstem neurons 3 days after injection with pHF (2.6 10⁵ ivu). Scale bar represents 30 m.

Full figure and legend (27K)

Alteration of frataxin levels in murine neurons by HSV-1 amplicon vectors

To characterize transgene expression at the biochemical level we infected mouse primary cortical neurons with the packaged HSV-1 amplicon vectors pHLC and pHF (Figure 2a and b). Using tubulin and actin as controls for loading quantity (Figure 2a, second and third panels), western blots of cell extracts from neuronal cultures from loxP[frda] mice infected with increasing amounts of pHLC vector revealed increasing amounts of -galactosidase and CRE recombinase expression (Figure 2a, first and fourth panels). The correlated decreasing amounts of frataxin in the same extracts (Figure 2a, bottom panel) indicate that the expressed CRE recombinase was clearly functional and able to inactivate the frda gene in the infected neurons by excising the floxed exon 4 (see ref. 18 and Figure 2c). We next confirmed our ability to increase frataxin expression using the pHF vector (Figure 2b). Using species-specific primers we detected human frataxin messenger RNA expression in neurons infected with the pHF vector but not in uninfected controls (Figure 2b, top panel) or neurons infected with pHSVlac (data not shown). Vector infection of neuronal cultures did not affect endogenous mouse frataxin expression (Figure 2b, second panel) with respect to the control glyceraldehyde phosphate dehydrogenase messenger RNA (Figure 2b, third panel). Using tubulin as a control for loading quantity (Figure 2b, fourth panel), western blot analysis of extracts from pHF-infected neurons showed that the total amount of frataxin was increased compared with control extracts (Figure 2b, bottom panel).

Figure 2.

Biochemical analysis of neuronal frda and frataxin alterations provoked by pHLC and pHF vectors. (a) Western blot analysis of primary cortical neurons from loxP[frda] mice infected with increasing volumes (0–50 l) of pHLC. Cell extracts were made 6 days after infection and western blots probed with antibodies against -galactosidase (-gal), -tubulin (tubulin), actin, CRE recombinase (CRE), and frataxin (1,250 polyclonal antibody). Increasing CRE dose results in a clear reduction in endogenous neuronal frataxin expression (bottom panel). (b) Reverse-transcriptase polymerase chain reaction (RT-PCR) and western blot analysis of primary cortical neurons from wild-type mice with no treatment (con) or 6 days after infection with pHF. RT-PCR of total cellular RNA (top three panels) using primers specific for human frataxin (hFRDA), mouse frataxin (mFrda), or glyceraldehyde 3-phosphate dehydrogenase (GAPDH) transcripts showed specific expression of human frataxin messenger RNA in neurons infected by pHF. Western blots of protein extracts (bottom two panels) probed with antibodies against tubulin and frataxin (1,250 polyclonal antibody) show a modest increase of frataxin levels in neurons infected by pHF. (c) Schematic diagram showing the location of the primers used to detect frda gene excision in loxP[frda] mice. Exons in the mouse frda gene are represented by numbered gray boxes. The gene conferring neomycin resistance is shown as the unshaded box labeled "neo," and loxP sites are indicated by arrowed boxes. Excision by CRE recombinase (CRE) converts the L3 allele (top) to the allele (bottom). The oligonucleotide primers P3, M2, and P5 hybridize at the locations indicated by the arrowheads, such that a 200-bp PCR product using P3 and P5 is obtained only in the presence of the allele. (d) PCR analysis showing neuronal frda gene excision 4 days after injection of pHLC into the brainstem. Genomic DNA was prepared from mice injected with 1.4 10⁴ infectious vector units (ivu) of pHSVlac (lane 1), 1.4 10⁴ ivu of pHLC (lanes 2 and 3), or 2 10⁵ ivu of pHLC (lane 4) and subjected to PCR using the L3 allele–specific primer pair M2 and P5 (top panel) or the allele–specific primer pair P3 and P5 (bottom panel). HSV, herpes simplex virus.

Full figure and legend (23K)

CRE-mediated elimination of frda in brainstem of loxP[frda] mice provokes motor coordination deficit

In FA patients, in addition to the spinal cord, degenerative changes are noted in the brainstem and the cerebellum. In trying to develop an animal model with high sensitivity to therapeutic effects, we reasoned that modulation of a small pool of key neurons by the focal administration of viral vectors in the brain might be sufficient to provoke an ataxic phenotype. We next injected the pHLC vector into the medulla of loxP[frda] mice and confirmed CRE recombinase function in vivoby polymerase chain reaction (PCR) (Figure 2d) using primers P3 and P5 specifically to detect the deleted frda allele (Figure 2c): the PCR product specific for the allele could be detected from brainstem genomic DNA from loxP[frda] mice injected with either low (Figure 2d, lanes 2 and 3) or high (Figure 2d, lane 4) vector concentrations of pHLC but not from genomic DNA of those injected with pHSVlac (Figure 2d, lane 1). Because genomic DNA was prepared from a brainstem sample including the injection site, the proportion of pHLC-infected cells in the total tissue mass dissected out would be expected to be quite low. Indeed, the PCR product specific for the undeleted L3 allele (Figure 2d, top panel) was clearly detectable and showed no change whether low (Figure 2d, lanes 2 and 3) or high doses (Figure 2d, lane 4) of pHLC were used, as the majority of the neurons in the dissected sample were not infected by the viral vector.

In mouse models, a well-established method of measuring motor coordination is the rotarod apparatus. To examine the phenotypic effect of frda gene excision we measured rotarod performance before treatment with viral vectors and at 2-week intervals after injection of pHLC and pHSVlac vectors (Figure 3) in loxP[frda] mice (n = 3) and wild-type C57/BL6 mice (n = 4). After 4 and 6 weeks, separation between the mean values of the pHLC- and the pHSVlac-injected loxP[frda] mice could be observed (Figure 3, top row), although the differences between these groups did not quite reach significance (P = 0.057 in both cases). After 8 weeks the rotarod performance difference between the two groups was significant (P = 0.03). This motor coordination deficit was clearly due to frda gene excision, as the difference between the two vectors was not observed when they were injected into wild-type C57/BL6 mice (Figure 3, second row). Interference of vector injection per se in rotarod performance was not significant, as shown by comparison of untreated mice with those injected with pHLC and pHSVlac (Figure 3, open circles in second row).

Figure 3.

Decreased rotarod performance of loxP[frda] mice injected with pHLC vector. Mice were analyzed in the accelerating rotarod test before treatment (week 0) and at 2-week intervals after stereotaxic injection in the brainstem of 3 l containing either pHSVlac vector (4.5 10⁶ infectious vector units/ml) or pHLC vector (4.7 10⁶ infectious vector units/ml). Mean values are shown with standard errors of the means (vertical bars). For transgenic loxP[frda] mice (top row), the group size was n = 3, and for wild-type mice (second row) the group size was n = 4. A third group of unoperated mice (n = 8, open circles, dotted lines) was included in the wild-type control experiments to ascertain the degree of interference of the surgical intervention in rotarod performance.

Full figure and legend (22K)

Motor coordination deficit provoked by frda gene excision can be reversed by frda cDNA delivery

Because a worsening in rotarod performance could be noted as early as 4 weeks after injection of CRE-expressing vector into loxP[frda] mice, we next investigated whether this motor deficit could be reversed by frda gene delivery with the pHF vector 4 weeks after the frda gene excision (Figure 4). We compared four groups of mice: those initially injected with pHLC and then with pHF ("rescued" group, n = 6, circles with dotted lines); those initially injected with pHLC and then with pHSVlac ("sham rescued" group, n = 5, circles with solid lines); those only injected initially with pHLC ("lesioned" group, n = 4, squares with solid lines); those only injected initially with pHSVlac ("sham lesioned" group, n = 5, squares with dotted lines). As in the previous experiment (Figure 3), by 4 weeks, all the mice initially injected with the pHLC vector worsened in rotarod performance compared with those injected with pHSVlac. Notably, by 4 weeks after the second injection (week 8) the mean values of the rescued group had reverted to those of the sham-lesioned group and were separated from the mean values of the lesioned group (Figure 4). The mean values of the sham rescued group were indistinguishable from those of the lesioned group, indicating that the second injection event had no non-specific rescue effect. This pattern of differences between the four groups was maintained throughout the remainder of the experiment (12 weeks after the second injection, week 16), with significant differences being observed between rescued and lesioned mice (P = 0.03) at the last time point measured (Figure 4). These results indicate that the neurological defect provoked by elimination of the frda gene in the brainstem/cerebellum can be reversed by re-expressing frda at a later time point. In this study the delay between lesion and rescue was 1 month, the minimum period required to observe a decrease in rotarod performance.

Figure 4.

Recuperation of rotarod performance in CRE-lesioned loxP[frda] mice injected with pHF vector. Mice were analyzed in the accelerating rotarod test before treatment (week 0) and at 2-week intervals after stereotaxic injection in the brainstem of 3 l containing either pHSVlac vector (4.5 10⁶ infectious vector units/ml) or pHLC vector (4.7 10⁶ infectious vector units/ml). Immediately after the rotarod test in week 4, mice were re-injected in the brainstem with 3 l containing either pHSVlac vector (4.5 10⁶ infectious vector units/ml) or pHF vector (4.8 10⁶ infectious vector units/ml). Mean values are shown with standard errors of the means (vertical bars). Double-headed arrows linking symbols on the right of the graphs for weeks 6, 10, and 16 indicate levels of significance between the indicated groups: dotted line, 0.05 < P < 0.1; solid line P < 0.05.

Full figure and legend (24K)

Discussion

In this study our main goal was to generate proof-of-principle data for frataxin gene therapy in the treatment of FA neuropathology. As animal models, we have generated localized frataxin knockout mice by targeting the inferior olive of transgenic loxP[frda] mice with CRE recombinase viral vectors. The chief advantage of our model is that the adult mouse brain contains only 30,000 olivary neurons^22,23 and histological analyses of various ataxic mutant mice show olivary neuronal deficits of between 20 and 60%.^{23,24,25,26,27} Thus our model should be sensitive to relatively low vector dosage, in the range of tens of thousands of infectious particles, permitting the use of unconcentrated HSV-1 amplicon stocks, which we have previously found to exhibit negligible toxicity in this anatomical area.¹⁶ Indeed, our results show that at these vector concentrations in vivo, changes in motor performance can be more readily detected than transgene expression, which required tenfold higher concentrations. This was due only to the limit of detection sensitivity, as we were able to detect CRE recombinase activity in vivo even at low vector concentrations.

Here we used frda gene excision in olivary neurons to provoke a deficit in rotarod performance that was stable over the 4-month period of the experiment. We also showed recovery from this deficit in motor coordination by restoration of frataxin expression using a HSV-1 amplicon vector. This functional recovery was surprisingly complete, in view of the fact that frataxin restoration can occur only in the overlapping subset of neurons transduced by both the first and second injections. Several explanations are possible. First, frda gene rescue of one subset of neurons may aid in the survival of others, by stimulating the generation of trophic factors and/or connections. Second, similar results were obtained by Fernandez et al.,²⁸ who used an ataxic rat model in which they provoked ablation of more than 80% of the inferior olive with 3-acetylpyridine. In this study, delayed insulin-like growth factor 1 treatment resulted in the survival of only 27% of the total olivary neurons but, remarkably, total functional recovery was observed. These authors attribute functional recovery not only to increased neuronal survival but also to induced axonal sprouting. Third, when recovery of motor coordination reaches a certain threshold, other compensatory mechanisms of learning and/or motor function may be sufficient to achieve a normal level of performance in given tasks such as the rotarod test.

The level of frataxin protein expressed by our pHF vector was similar to the endogenous levels observed in wild-type neurons (compare the left and right lanes in the bottom panel of Figure 1d), indicating that we had obtained physiologically relevant levels of transgene expression. This may be particularly important, because high-level overexpression of frataxin by lentiviral or adeno-associated viral vectors has been previously reported to be toxic in FA patient fibroblasts.¹³ We have recently demonstrated rescue of the FA fibroblast phenotype with negligible toxicity by using HSV-1 amplicon vectors bearing the entire FRDA locus,²⁹ which results in physiological expression levels. It is also known that frataxin expression is reduced to 40% in asymptomatic human carriers of pathogenic FRDA mutations,³⁰ indicating that high-level frataxin expression is not required for therapeutic recovery. In this work, despite the fact that expression from viral promoters has been observed to diminish significantly in neurons in vivo, we have used both the cytomegalovirus and the HSV-1 IE4/5 promoters to drive transgene expression for several reasons: first, we needed only transient expression of CRE recombinase (driven by the cytomegalovirus promoter) to effect frda excision in loxP[frda] neurons; second, in contrast to other brain areas,³¹ we have previously observed that in the inferior olive and cerebellum, expression from the IE4/5 promoter is stable for at least 40 days;¹⁶ third, for this proof-of-concept study, the simplest amplicon vectors were used for ease of construction. Our ongoing studies are focused on the use of our HSV-1 amplicon vectors bearing the whole FRDA genomic locus,²⁹ as the presence of the native regulatory sequences, including promoter, enhancer, and silencer elements, will likely produce physiological and stable expression.

All therapeutic strategies aimed at alleviating neurodegeneration in FA and many other neurological conditions have failed so far, emphasizing the need to explore other options such as gene therapy. Transgene delivery to the nervous system is by no means an easy task; introduction of vectors into the bloodstream is generally an ineffective way to target them into neurons, and most methods rely upon direct injection into nervous tissue, which in turn results in transduction predominantly in the area near the injection site. HSV-1 is a promising tool for nervous system gene therapy³² for a number of reasons: it naturally targets neurons for infection and can be transported throughout the nervous system to establish lifelong persistence; the virion has an extremely large packaging capacity; and the genome remains episomal in infected cells, thus avoiding a risk of insertional mutagenesis. Here we have described the functional validation of HSV-1 viral vectors by delivery into the brainstem using stereotaxic injection to achieve focal gene deletion and rescue in neurons that affect motor coordination. During the preparation of this manuscript new transgenic knock-in mouse FA models were reported that closely mimic the human disease because they have FRDA transgenes that bear pathological GAA repeat expansions.³³ These whole-animal models will greatly facilitate our ongoing studies, which are aimed at identifying the means to achieve suitable penetration of our therapeutic vectors into FA-relevant neuronal targets.

Materials and Methods

Reagents and plasmids. The polyclonal antiserum directed against calbindin (1:500) was from Chemicon (Temecula, CA), as was the frataxin monoclonal antibody 1G2 (1:100). The 1,250 anti-frataxin polyclonal antiserum (1:1,000) was a kind gift from Helene Puccio. Both frataxin antibodies recognize both the mouse and human proteins, but in our hands only the 1G2 antibody was suitable for immunocytochemistry and only the 1,250 antiserum was suitable for western blot analysis. Anti-FLAG polyclonal antiserum (1:5,000) was a kind gift from Rachael L. Neve, and the polyclonal antibody against -galactosidase (1:5,000) was from MP Biomedicals (Eschwege, Germany). Monoclonal antibodies specific for -tubulin (1:1,000) and -actin (1:1,000) were from Sigma (Madrid, Spain),and the polyclonal antibody against CRE recombinase (1:1,000) was from Abcam (Cambridge, UK). FLAG immunochemistry was performed with peroxidase-conjugated secondary antibody (Amersham, Buckinghamshire, UK) using the substrate 3,3'-diaminobenzidine (Sigma, Madrid, Spain). Calbindin and -galactosidase immunofluorescence was performed using Alexa 488–conjugated secondary antibodies (Invitrogen, Barcelona, Spain). The pHLC plasmid was generated by inserting the CRE recombinase cDNA driven by the cytomegalovirus immediate early promoter as an AseI/XbaI fragment derived from pCI-cre (a kind gift from Fabio Rossi) into pHSVlac.²¹ We generated the transgene encoding the frataxin-FLAG fusion protein by PCR amplification from a human frda cDNA clone in pCR-Script-SK+ (a kind gift from Francesc Palau and Jose Gonzalez Castaño) using the T7 primer at the 5' end and the oligonucleotide 5'-GTCACTAGTGAATTCACGATTTATCGTCATCGTCTTTGTA
GTCCATAGCATCTTTTCCGGAATAGG-3' at the 3' end. The PCR product was then cloned as a BamHI/EcoRI fragment into pHSVpuc³⁴ to generate pHF. Reverse-transcriptase PCR analysis of human FRDA expression in mouse primary cortical neurons was performed as previously described.²⁹

Viral vectors. HSV-1 amplicon vectors were packaged using as helper virus the ICP27 deletion mutant 5dl1.2 virus³⁵ grown on the complementing 2-2 cell line.³⁶ Vector packaging was performed as previously described.¹⁴ For high-dose experiments, HSV-1 amplicons were concentrated by centrifuging viral preparations through a 25% sucrose cushion at 100,000g for 3 hours and re-suspending the pellet in phosphate-buffered saline (PBS) overnight at 4°C. Titering of infectious vector units (ivu) was performed by Xgal staining (pHSVlac and pHLC) or FLAG immunostaining (pHF) in Vero cells, and titering of helper virus plaque-forming units (pfu) in 2-2 cells was performed as previously described.¹⁴ Titers of unconcentrated HSV-1 amplicon stocks were 4.5 10⁶ ivu/ml and 1.5 10⁶ pfu/ml for pHSVlac, 4.7 10⁶ ivu/ml and 5 10⁶ pfu/ml for pHLC, and 4.8 10⁶ ivu/ml and 6 10⁶ pfu/ml for pHF. Titers of the concentrated amplicon stocks were 6.7 10⁷ ivu/ml and 8.0 10⁷ pfu/ml for pHLC and 8.7 10⁷ ivu/ml and 7.0 10⁷ pfu/ml for pHF. No wild-type HSV-1 revertants were found in any of the stocks within the sensitivity of our pfu assay in Vero cells (10^-6). HSV-1 amplicon vectors were packaged and used to infect cells in culture as well as adult mice (intracerebral stereotaxic injection) in P2 facilities. Laboratory facilities at the Centro de Biología Molecular are subject to regular inspection by government authorities and comply with the following national regulations: Royal Decree 664/1997 (Protection of workers against risks related with exposure to biological agents during work); Royal Decree 951/1997 (requirement for legal approval of the confined use, deliberate release and commercialisation of genetically modified organisms, with the aim of preventing risks to human health and the environment).

Neuronal cultures. Cortical primary neurons were cultured according to an established protocol.^37,38 In brief, 17-day-old embryos were removed from a pregnant female mouse and dissected in pre-chilled Hank's buffered salt solution. After removal of the meninges, striatum, and hippocampus, the intact cortices were cut into small pieces and then incubated in a 0.25% trypsin (Sigma, Madrid, Spain), 1 mg/ml DNaseI (Roche, Barcelona, Spain) solution in Hank's buffered salt solution without calcium and magnesium for 15 minutes at 37°C, with gentle shaking every 3–4 minutes. The dissociated cells were then re-suspended and plated in minimum essential medium supplemented with 0.6% glucose and 10% horse serum (Invitrogen, Barcelona, Spain). The medium was changed after 4 hours to Neurobasal culture medium supplemented with B-27, 2 mmol GlutaMaxI (both from Invitrogen, Barcelona, Spain), and a mix of penicillin and streptomycin (100 U/ml and 100 g/ml, respectively). The cells were plated at a density of 1 10⁵ cells/cm² onto poly-l-lysine-precoated surfaces, and one third of the medium was replaced every 3 days. To avoid glial proliferation, 5 mol Ara-C was added 24 hours after plating.

Western blot analysis. For protein extracts, cells were washed once with PBS, placed on ice, and then homogenized in a buffer containing 20 mmol HEPES, pH 7.4; 100 mmol sodium chloride (NaCl); 100 mmol sodium fluoride; 1% Triton X-100; 1 mmol sodium orthovanadate; 5 mmol EDTA; and the COMPLETE protease inhibitor cocktail (Roche, Barcelona, Spain). After determination of the protein content using the Bradford assay, samples containing the same amount of protein were mixed with electrophoresis buffer containing sodium dodecylsulfate, boiled for 5 minutes, and separated by gel electrophoresis in the presence of sodium dodecylsulfate on 6–10% acrylamide gels. The proteins were then transferred to nitrocellulose membranes following standard procedures and the membranes were blocked with 10% non-fat dried milk in PBS, 0.2% Tween-20. The blocked membranes were incubated overnight with primary antibodies diluted in blocking solution at 4°C. The filters were then washed three times in PBS, 0.2% Tween-20 and incubated with the corresponding peroxidase-conjugated secondary antibody for 1 hour at room temperature. After three further washes with PBS the immunoreactive proteins were visualized using an enhanced chemiluminescence detection system (Amersham, Buckinghamshire, UK).

Mouse maintenance. Transgenic mice hemizygous for the conditional frataxin allele (Frda+^/L3)¹⁸ were kindly provided by Helene Puccio, Michel Koenig, and Ignacio Torres. These were bred in a C57/BL6 genetic background to generate homozygous Frda^L3/L3 mice (referred to in the text as loxP[frda] mice) using the following PCR primers for genotyping: P2neo, P3,¹⁸ and MOF1: 5'-AACGTTACTCTTAGGGTCAG-3'. The primer pair P3 + P2neo generates an 250–base pair PCR product specific for the presence of the L3 allele, and the primer pair P3 + MOF1 generates an 350–base pair PCR product specific for the wild-type allele. Mice were cared for by trained personnel in the Centro de Biología Molecular animal house, with restricted access. The experimental protocol was approved by the Institutional Animal Care and Use Committee at the Centro de Biología Molecular and followed European Directive 86/609/CEE and its posterior modification 2003/65/CE as implemented by Spanish law 1201/2005. All efforts were made to minimize suffering and the number of animals used.

Rotarod test for motor coordination. Motor coordination was determined in a rotarod apparatus (Ugo Basile, Italy). Animals were tested using the accelerating rotating rod protocol, from 4 to 40 rpm over 5 minutes followed by 5 minutes at maximum velocity. Latency was defined as the duration for which the mice were able to maintain their equilibrium on the accelerating rod until falling off, and was recorded in four consecutive trials. In the experiments shown in Figure 3, the four trials were conducted in the mornings and afternoons over 2 days with a minimum of 4 hours between trials. In the experiments shown in Figure 4, trials were conducted on consecutive days. Mice that were still on the rod at the end of the trial were scored with a latency of 10 minutes. One week before the first measurement, the mice received a training session to familiarize them with the procedure. Results were analyzed using a repeated measure analysis of variance test with three factors: trial number (fixed); vector (fixed); animals (variable), nested in vector and crossed with trial number.

Stereotaxic injection. Mice 3–6 months of age were anesthetized using isofluorane and when they no longer demonstrated the footpad pinch reflex, they were positioned in a Stoelting Lab StandardJ stereotaxic instrument fitted with a mouse gas mask (#51609; Stoelting, Wood Dale, IL) and a burr hole was drilled into the skull at the midline, 2.7 mm rostral from lambda. A 26-gauge needle with 30° bevel attached to a 5-l Hamilton syringe was aligned over the burr hole and then lowered into the brain 5.5 mm deep from the surface of the skull. Injections of 3 l HSV-1 amplicon vector or Fluoro-ruby (Molecular Probes, Eugene, OR) were performed over 5 minutes, after which the needle was withdrawn slowly over 3 minutes.

After surgery, the incision was closed with sutures and the animals were allowed to recover on a heated pad before being returned to their cages. For histochemical analyses brain tissues were fixed by perfusion with cold paraformaldehyde (4% in PBS), cryoprotected by overnight incubation in 30% sucrose, and sectioned at -20°C in 30-m slices after embedding in Tissue Tek (Sakura Finetek Europa, Zoeterwoude, The Netherlands).

References

REFERENCES

1. Campuzano, V, Montermini, L, Molto, MD, Pianese, L, Cossee, M and Cavalcanti, F (1996). Friedreich's ataxia: autosomal recessive disease caused by an intronic GAA triplet repeat expansion. Science 271: 1423–1427. | Article | PubMed | ISI | ChemPort |
2. Pianese, L, Tammaro, A, Turano, M, De Biase, I, Monticelli, A and Cocozza, S (2002). Identification of a novel transcript of X25, the human gene involved in Friedreich's ataxia. Neurosci Lett 320: 137–140. | Article | PubMed | ChemPort |
3. Bartolo, C, Mendell, JR and Prior, TW (1998). Identification of a missense mutation in a Friedreich's ataxia patient: implications for diagnosis and carrier studies. Am J Med Genet 79: 396–399. | Article | PubMed | ChemPort |
4. Bidichandani, SI, Ashizawa, T and Patel, PI (1997). Atypical Friedreich ataxia caused by compound heterozygosity for a novel missense mutation and the GAA triplet-repeat expansion. Am J Hum Genet 60: 1251–1256. | PubMed | ChemPort |
5. Cossee, M, Durr, A, Schmitt, M, Dahl, N, Trouillas, P, Allinson, P et al. (1999). Friedreich's ataxia: point mutations and clinical presentation of compound heterozygotes. Ann Neurol 45: 200–206. | Article | PubMed | ISI | ChemPort |
6. McCormack, ML, Guttmann, RP, Schumann, M, Farmer, JM, Stolle, CA, Campuzano, V et al. (2000). Frataxin point mutations in two patients with Friedreich's ataxia and unusual clinical features. J Neurol Neurosurg Psychiatry 68: 661–664. | Article | PubMed | ChemPort |
7. Campuzano, V, Montermini, L, Lutz, Y, Cova, L, Hindelang, C, Jiralerspong, S et al. (1997). Frataxin is reduced in Friedreich ataxia patients and is associated with mitochondrial membranes. Hum Mol Genet 6: 1771–1780. | Article | PubMed | ISI | ChemPort |
8. Miranda, CJ, Santos, MM, Ohshima, K, Tessaro, M, Sequeiros, J and Pandolfo, M (2004). Frataxin overexpressing mice. FEBS Lett 572: 281–288. | Article | PubMed | ChemPort |
9. Pook, MA, Al-Mahdawi, S, Carroll, CJ, Cossee, M, Puccio, H, Lawrence, L et al. (2001). Rescue of the Friedreich's ataxia knockout mouse by human YAC transgenesis. Neurogenetics 3: 185–193. | PubMed | ChemPort |
10. Sarsero, JP, Li, L, Holloway, TP, Voullaire, L, Gazeas, S, Fowler, KJ et al. (2004). Human BAC-mediated rescue of the Friedreich ataxia knockout mutation in transgenic mice. Mamm Genome 15: 370–382. | Article | PubMed | ChemPort |
11. Friedreich, N (1876). Ueber Ataxia mit besonderer Berucksichtigung der hereditaren Formen. Virchow's Arch Pathol Anat 68: 145–245. | Article |
12. Davidson, BL and Breakefield, XO (2003). Viral vectors for gene delivery to the nervous system. Nat Rev Neurosci 4: 353–364. | Article | PubMed | ISI | ChemPort |
13. Fleming, J, Spinoulas, A, Zheng, M, Cunningham, SC, Ginn, SL, McQuilty, RC et al. (2005). Partial correction of sensitivity to oxidant stress in Friedreich ataxia patient fibroblasts by frataxin-encoding adeno-associated virus and lentivirus vectors. Hum Gene Ther 16: 947–956. | Article | PubMed | ChemPort |
14. Lim, F, Hartley, D, Starr, P, Lang, P, Song, S, Yu, L et al. (1996). Generation of high-titer defective HSV-1 vectors using an IE 2 deletion mutant and quantitative study of expression in cultured cortical cells Biotechniques 20: 460–469. | PubMed | ChemPort |
15. Neve, RL and Lim, F (1999). Overview of gene delivery into cells using HSV-1-based vectors. In: Crooks, G (ed.) Current Protocols in Neuroscience Greene Publishing Associates and Wiley-Interscience, New York. pp. 4.12.11–14.12.17.
16. Agudo, M, Trejo, JL, Lim, F, Avila, J, Torres-Aleman, I, Diaz-Nido, J et al. (2002). Highly efficient and specific gene transfer to Purkinje cells in vivousing a herpes simplex virus I amplicon. Hum Gene Ther 13: 665–674. | Article | PubMed | ChemPort |
17. Llinas, R, Walton, K, Hillman, DE and Sotelo, C (1975). Inferior olive: its role in motor learing. Science 190: 1230–1231. | Article | PubMed | ISI | ChemPort |
18. Puccio, H, Simon, D, Cossee, M, Criqui-Filipe, P, Tiziano, F, Melki, J et al. (2001). Mouse models for Friedreich ataxia exhibit cardiomyopathy, sensory nerve defect and Fe-S enzyme deficiency followed by intramitochondrial iron deposits. Nat Genet 27: 181–186. | Article | PubMed | ISI | ChemPort |
19. Rondi-Reig, L, Delhaye-Bouchaud, N, Mariani, J and Caston, J (1997). Role of the inferior olivary complex in motor skills and motor learning in the adult rat. Neuroscience 77: 955–963. | Article | PubMed | ChemPort |
20. Libby, RT, Cosman, D, Cooney, MK, Merriam, JE, March, CJ and Hopp, TP (1988). Human rhinovirus 3C protease: cloning and expression of an active form in Escherichia coli. Biochemistry 27: 6262–6268. | Article | PubMed | ChemPort |
21. Geller, AI and Breakefield, XO (1988). A defective HSV-1 vector expresses Escherichia coli beta-galactosidase in cultured peripheral neurons. Science 241: 1667–1669. | Article | PubMed | ISI | ChemPort |
22. Zanjani, HS, Vogel, MW, Martinou, JC, Delhaye-Bouchaud, N and Mariani, J (1998). Postnatal expression of Hu-bcl-2 gene in Lurcher mutant mice fails to rescue Purkinje cells but protects inferior olivary neurons from target-related cell death. J Neurosci 18: 319–327. | PubMed | ChemPort |
23. Blatt, GJ and Eisenman, LM (1985). A qualitative and quantitative light microscopic study of the inferior olivary complex of normal, reeler, and weaver mutant mice J Comp Neurol 232: 117–128. | Article | PubMed | ChemPort |
24. Caddy, KW and Biscoe, TJ (1979). Structural and quantitative studies on the normal C3H and Lurcher mutant mouse. Philos Trans R Soc Lond B Biol Sci 287: 167–201. | PubMed | ChemPort |
25. Blatt, GJ and Eisenman, LM (1985). A qualitative and quantitative light microscopic study of the inferior olivary complex in the adult staggerer mutant mouse. J Neurogenet 2: 51–66. | PubMed | ChemPort |
26. Shojaeian, H, Delhaye-Bouchaud, N and Mariani, J (1985). Decreased number of cells in the inferior olivary nucleus of the developing staggerer mouse. Brain Res 353: 141–146. | PubMed | ChemPort |
27. Shojaeian, H, Delhaye-Bouchaud, N and Mariani, J (1985). Neuronal death and synapse elimination in the olivocerebellar system. II. Cell counts in the inferior olive of adult x-irradiated rats and weaver and reeler mutant mice. J Comp Neurol 232: 309–318. | Article | PubMed | ChemPort |
28. Fernandez, AM, de la Vega, AG and Torres-Aleman, I (1998). Insulin-like growth factor I restores motor coordination in a rat model of cerebellar ataxia. Proc Natl Acad Sci USA 95: 1253–1258. | Article | PubMed | ChemPort |
29. Gomez-Sebastian, S, Gimenez-Cassina, A, Diaz-Nido, J, Lim, F and Wade-Martins, R (2007). Infectious delivery and expression of a 135 kb human FRDA genomic lous complements Friedreich's ataxia deficiency in human cells. Mol Ther 15: 248–254. | Article | PubMed |
30. Pianese, L, Turano, M, Lo Casale, MS, De Biase, I, Giacchetti, M, Monticelli, A et al. (2004). Real time PCR quantification of frataxin mRNA in the peripheral blood leucocytes of Friedreich ataxia patients and carriers. J Neurol Neurosurg Psychiatry 75: 1061–1063. | Article | PubMed | ChemPort |
31. Starr, PA, Lim, F, Grant, FD, Trask, L, Lang, P, Yu, L et al. (1996). Long-term persistence of defective HSV-1 vectors in the rat brain is demonstrated by reactivation of vector gene expression. Gene Ther 3: 615–623. | PubMed | ChemPort |
32. Berges, BK, Wolfe, JH and Fraser, NW (2007). Transduction of brain by herpes simplex virus vectors. Mol Ther 15: 20–29. | Article | PubMed |
33. Al-Mahdawi, S, Pinto, RM, Varshney, D, Lawrence, L, Lowrie, MB, Hughes, S et al. (2006). GAA repeat expansion mutation mouse models of Friedreich ataxia exhibit oxidative stress leading to progressive neuronal and cardiac pathology. Genomics 88: 580–590. | PubMed | ChemPort |
34. Geller, AI, During, MJ, Haycock, JW, Freese, A and Neve, R (1993). Long-term increases in neurotransmitter release from neuronal cells expressing a constitutively active adenylate cyclase from a herpes simplex virus type 1 vector. Proc Natl Acad Sci USA 90: 7603–7607. | Article | PubMed | ChemPort |
35. McCarthy, AM, McMahan, L and Schaffer, PA (1989). Herpes simplex virus type 1 ICP27 deletion mutants exhibit altered patterns of transcription and are DNA deficient. J Virol 63: 18–27. | PubMed | ISI | ChemPort |
36. Smith, IL, Hardwicke, MA and Sandri-Goldin, RM (1992). Evidence that the herpes simplex virus immediate early protein ICP27 acts post-transcriptionally during infection to regulate gene expression. Virology 186: 74–86. | Article | PubMed | ISI | ChemPort |
37. Bartlett, WP and Banker, GA (1984). An electron microscopic study of the development of axons and dendrites by hippocampal neurons in culture. I. Cells which develop without intercellular contacts. J Neurosci 4: 1944–1953. | PubMed | ISI | ChemPort |
38. Banker, G and Goslin, K (1998). Culturing nerve cells. MIT Press: Cambridge, MA.

Acknowledgments

This work was supported by grants from the Friedreich's Ataxia Research Alliance, the Caixa Foundation, and the Spanish ministries of Education and Science and of Health. F.L. is the recipient of a Ramon y Cajal contract from the Spanish Ministry of Education and Science. G.M.P. and A.G.-C. were supported by Formación de Profesorado Universitario fellowships from the Spanish Ministry of Education and Science. B.I. was the recipient of a European Molecular Biology Organizaton re-incorporation fellowship. We are grateful to Rachael Neve (McLean Hospital, Boston) for the anti-FLAG antiserum; Fabio Rossi (University of British Columbia) for the plasmid pCI-cre; Francesc Palau (Instituto de Biomedicina, Valencia) and Jose Gonzalez Castaño (Universidad Autónoma de Madrid) for the frataxin cDNA; Helene Puccio (Institut de Genetique et de Biologie Moleculaire et Cellulaire, Illkrich, France) for the anti-frataxin polyclonal antibody and, together with Michel Koenig (Institut de Genetique et de Biologie Moleculaire et Cellulaire, Illkirch, France) and Ignacio Torres (Instituto Cajal, Madrid), for the Frda+^/L3 mice. At the Centro de Biología Molecular "Severo Ochoa," Madrid, we thank Jesus Avila for his support, Jose J. Lucas for the loan of the rotarod apparatus, Javier Palacín, all the animal house staff, and the Microscopy Service.