(A) We determined frataxin mRNA levels by quantitative real-time RT-PCR in brain, heart and cerebellum of 8 weeks old WT and KIKI mice (n = 8). All values are normalized to RER1 and β2m mRNA levels and are relative to WT brain, set to 100. Error bars are SD. Data were generated in two independent experiments. (** P<0.001). (B) Levels of H4K5, H4K8, H4K16, H3K14 acetylation and of H3K9 trimethylation in KIKI and WT mice brains (n = 8). We performed ChIP experiments with antibodies for murine histones H3 and H4 that carry each of the indicated modifications. We used primer pairs from the first intron of the mouse frataxin gene just upstream of the point of insertion of the GAA repeat in KIKI mice. Relative recovery, determined by quantitative real-time RT-PCR, is expressed in relation to GAPDH and the recovery in samples from WT animals is set to 100 for each antibody. Error bars are SEM. of four independent immunoprecipitations, each quantified in triplicate. (C) Histogram showing the number of genes differentially expressed in KIKI mouse brain (cerebral hemispheres), cerebellum and heart compared to WT mice. Genes with increased expression in KIKI mice are in red, genes with decresed expression in KIKI mice are in green. Only genes showing changes at P<0.005 are counted.

(A) Structure of HDACI 106. (B) Histone H4 acetylation in normal fibroblast cells (GM08333 Coriell Cell Repositories, Camden, NJ) was measured by Western blot of total cell lysate after 24 hr incubation with indicated compounds. 9b (N¹-(2-methoxyphenyl)-N⁷-phenylheptanediamide) has been described [22]. Compounds were dissolved in DMSO prior to addition to cell culture to give the final concentration indicated above at 0.1% DMSO. (C) FXN mRNA levels were measured in RNA from isolated PBMCs from carrier/patient pairs (1–3) after a 48 hour incubation with 106 or DMSO (at 0.1%). Quantitative real-time RT-PCR was used to determine relative FXN levels between each carrier/patient pair using GAPDH as a control housekeeping gene. Error is represented as the standard deviation of the mean from 3 determinations. Primary lymphocytes were obtained from donor blood from FRDA patients and their carrier relatives (under an approved Human Subjects Protocol, with appropriate informed consent) as described [22].

(A) The effect 106 on histone acetylation persists for at least 24 hours in KIKI mice. Mice were sacrificed, then extraction of brain acid-soluble nuclear proteins, SDS-PAGE, and western blotting with antibodies to total histone H3 and to acetylated histone H3 were performed. Sample in lane a is from the brain of a 106-treated mouse sacrificed after four hours, in lane b is from the brain of a vehicle-treated mouse, in lane c is from the brain of a 106-treated mouse sacrificed after 24 hours. (B) Frataxin mRNA levels in WT mice and in KIKI mice treated with either vehicle or HDAC inhibitor 106. Frataxin levels were determined by quantitative real-time RT-PCR relative to RER1 and β2m mRNAs, both unaffected by the HDACI. The WT frataxin mRNA level in each tissue was set to 100. Brain: n = 12. Cerebellum and Heart: n = 8. Data from 3 separate experiments were pooled and bars indicate SD. (** P<0.001; * P<0.05). (C) Western blot of frataxin in brain extracts of KIKI mice that had been treated with daily injections of 150 mg/Kg of 106 (lanes B5–B8) or vehicle (lanes B1–B4) for three days. Mice were sacrified 24 hours after the last injection. Actin was detected for normalization. (D) Densitometric analysis of the frataxin western blot shown in (C). (E) Levels of H4K5, H4K8, H4K16 and H3K14 acetylation of H3K9 trimethylation in KIKI mice treated with one subcutaneous injection per day of 150 mg/kg of 106 for 3 consecutive days compared to vehicle-treated KIKI littermates and WT (n = 6). We performed ChIP experiments with antibodies for murine histones H3 and H4 carrying each modification. Primer pairs corresponded to the first intron of the mouse frataxin gene just upstream of the point of insertion of the GAA repeat in KIKI mice. Relative recovery, determined by quantitative real-time PCR, is expressed in relation to GAPDH and the recovery in samples from WT animals is set to 100 for each antibody. Error bars are SEM of four independent immunoprecipitations, each quantified in triplicate.

(A) 670 probes (655 unique genes) in total were affected by drug treatment in either tissue in either strain. A few drug-responsive genes were shared between the two strains. Quantification of the number of genes (among the 615 KIKI-related) that go in opposite direction after drug treatment show that 70–85% genes change but only in KIKI samples. In a subset of genes (mostly in cerebellum) this difference was statistically significant at P<0.005. (B) Heatmaps representing the gene expression profiling before and after treatment with 106 in brain (left), cerebellum (center) and heart (right) of KIKI mice (four replicates per condition). Fold changes of the genes differentially expressed in KIKI vs. WT (brain = 61, cerebellum = 318, and heart = 182 genes at P<0.005) are depicted before treatment (KIKI vs. WT, grey bars on the top) and after treatment (KIKI treated vs. KIKI untreated, blue bars on the top). Shades of red represent upregulation, shades of green downregulation. Genes and samples are clustered by similarity. (C) Most genes (67% in brain, 84% in cerebellum, 67% in heart) show coordinate changes towards normal levels after treatment. In a subset of these genes (34% in brain, 23% in cerebellum, 60% in heart) these changes reach statistical significance.