Author Contributions
Conceived and designed the experiments: ADJ GLG GJR. Performed the experiments: ADJ WL IS BT. Analyzed the data: ADJ WL. Contributed reagents/materials/analysis tools: IS BT GLG. Wrote the paper: ADJ WL GJR.

060919 (B) cells. 020913 and 060919 cells were treated with FDA-approved RTK inhibitors at 10% of their IC50 concentration. Combination of gefitinib (2 mM) and sunitinib (4 mM)

cells of an analog of BML-210, HDACi 4b, resulted in a 2.5-fold enhancement of FXN mRNA (at 5 mM), acetylation of H3K14, H4K5 and H4K12 in the chromatin region immediately upstream of the GAA repeats, and a 3.5-fold increase in FXN protein levels (at 2.5 mM) [5]. A subsequent short pharmacodynamic study in a FRDA mouse model showed that a close analogue of HDACi 4b, the tolyl derivative compound 106, corrected the FXN deficiency [6]. These mice carry a homozygous (GAA)230 expansion in the first intron of the mouse FXN gene (KI/KI mice) [7]. Biochemical analysis revealed that these mice carry the same heterochromatin marks, close to the GAA repeat, as those detected in patient cell lines and have mildly but significantly reduced FXN mRNA and protein levels; however, they show no overt phenotype. Compound 106 given at 150 mg/kg subcutaneously once daily for 3 days increased global brain tissue histone acetylation as well as histone acetylation close to the GAA repeat and restored FXN levels in the nervous system and heart. Reversion of other differentially expressed genes towards wild type levels was also observed. Compound 106 showed no apparent toxicity in this study. Recently, the long-term benefit of chronic subcutaneous administration of three pimelic o-aminobenzamide inhibitors (compounds 106, 136 and 109) were assessed in another mouse model of FRDA. This mouse model (YG8R) contains the human FXN gene with expanded GAA repeats in a mouse FXN null background [8,9]. These mice show an approximate 30% reduction in FXN protein levels, mildly impaired motor coordination in females, reduced aconitase enzyme activity and DRG neuronal pathology, as well as a modest non-significant reduction in weight. However, YG8R mice show no evidence of hypoacetylation of H3 or H4 histones relative to WT or a reduction in FXN mRNA compared to WT [9]. The HDAC inhibitors were administered at 150 mg/kg (106), 50 mg/kg (136) and 100 mg/kg (109) by 3 (106) or 5 (136 and 109) subcutaneous injections per week to YG8R and WT mice for 4.5 to 5 months; the rationale for the different dosing and frequency were not given, and to our knowledge, no ADME data has been presented on this series. Although generally well tolerated, the inhibitors gave variable results. The authors concluded that prolonged treatment with any of the three HDAC inhibitors 106, 136 and 109 ameliorated FRDA disease-like pathology to some extent, and speculated that the apparent discrepancy in outcome with the three inhibitors could be due to differences in their potency, specificity, tissue distribution, and brain penetrance, as well as differences in dose levels and dose frequency resulting in sub maximal exposure [10].

the autophagy-lysosomal system, and impairment of synaptic transmission and plasticity. HDAC inhibition has been proposed as a therapeutic strategy for HD (reviewed in [14?6]). Indeed, broad-spectrum HDAC inhibitors partially rectify the transcriptional dysregulation in HD cell and animal models [17?3], enhance the degradation of mHTT by altering the acetylation state of key residues within the protein [24?7], and improve cognition through enhancement of learning and memory processes [28,29]. Thomas et al showed that HDACi 4b has a therapeutic effect in the R6/2 HD mouse model [30]. The R6/2 strain used in this study expresses the exon 1 HTT protein with an expanded polyglutamine region of ,300 repeats (R6/2300Q), and manifests a delayed phenotype compared to the better characterised R6/2 model that has a shorter polyglutamine expansion [31?4]. The R6/2300Q mice exhibit significant deficits in motor behaviour by 12 weeks of age, striatal atrophy, and survive 6 to 7 months. A short pharmacodynamic study (once daily subcutaneous treatment with 150 mg/kg 4b for 3 days) successfully ameliorated gene expression abnormalities in these mice and showed increased histone H3 acetylation in association with selected down-regulated genes. In a chronic efficacy study, 4b was complexed to 2hydroxypropyl-b-cyclodextrin and diluted in drinking water (estimated dosage of 150 mg/kg/day) and given to mice from 4 months of age. However, the expected differences in oral versus parenteral administration were not addressed in the Thomas et al. study [30]. While this precludes direct correlation between the pharmacodynamic studies and the results of the efficacy trial, these mice showed improved motor performance and overall appearance and an amelioration of body weight loss. Gross brain weight and striatal volume were also improved on termination of the study at 6 months of age. The successful use of 4b in treating R6/2 mice loosely correlates with an earlier report, in which the hydroxamic acid HDAC inhibitor SAHA was administered in drinking water to R6/2 mice that harbour the smaller polyglutamine repeat (,200 Q) and exhibit a more aggressive phenotype [22]. These animals also showed significant improvement in motor dysfunction as assessed by rotarod performance and grip strength, but this improvement was offset by the failure of both wild type and R6/2 mice to gain weight at the maximum tolerated dose (0.67 g/L in drinking water), suggestive of a narrow therapeutic window.

Pharmacology of Pimelic Diphenylamide-based HDAC Inhibitors
Subsequent reports provided an intriguing explanation for the efficacy and well-tolerated effects of the pimelic diphenylamidebased inhibitors. A common feature of these HDAC inhibitors (also known as ortho-N-acyl-phenylene diamines or benzamides) is an acylated ortho-phenylene diamine unit, which is thought to interact with the zinc ion of HDACs. Compounds from this series are selective for HDAC1, HDAC2, and HDAC3 over other HDAC isoforms, with no activity reported against HDAC Class IIa enzymes and only weak activity reported against HDAC8 [35,36]. Furthermore, they appear to bind to the catalytic site of these HDACs via a unique binding mode not shared with other hydroxamic acid based inhibitors; a time-dependent increase in affinity with an extremely slow off rate has been observed [36?8]. These exciting findings plus the report of in vivo efficacy and a neuroprotective profile in the R6/2 HD model prompted us to synthesize and evaluate 4b to independently validate this finding in the more widely used lower CAG repeat length R6/2 model [31?4], and also potentially in other HD rodent models. We had two objectives: to further validate the finding that preferential

Leave a Reply