First, we examined levels of class I HDAC proteins at different stages of disease progression in the transgenic R6/2 mouse model of HD. In R6/2 mice high expression levels of a mutant huntingtin exon I fragment containing ~160 CAG repeats lead to early neurological phenotype and premature death at 110-120 days 
We assessed expression levels of HDAC1 and HDAC3 proteins in cortex and striatum, the primary brain areas affected by HD, of R6/2 transgenic mice and wild-type littermates. HDAC profiling in cortex was performed on mice sacrificed at 4 and 12 weeks of age, and in striatal samples on animals sacrificed at 9 weeks of age. Increased levels of HDAC1 protein in cortical samples from R6/2 mice were observed at both time points (Fig. 1 A-C
). Similarly, in the striatum from 9 week-old R6/2 mice, levels of HDAC1 were elevated (Fig. 1 D, E
Figure 1. Levels of class I HDAC proteins in cortex and striatum of R6/2 HD mouse model. A) Western blot showing class I HDAC1 and 3 protein levels in cortices of 4 week-old R6/2 transgenic (R6/2) and wild-type (wt) mice. B) Western blot showing class (more ...)
Next, we repeated profiling of HDAC class I proteins in CAG140 full-length knock-in HD mice. The CAG140 knock-in mouse model is genetically more representative of the human disease and has analogous neuropathology, but these mice develop only a modest neurological phenotype during their lifespan 
. We analyzed levels of class I HDAC proteins in cortices from young (8 month-old) and aged (24 month-old) CAG140 knock-in mice and their wild-type littermates. The levels of HDAC1, HDAC2 and HDAC3 in CAG140 knock-in animals at 8 and at 24 months of age were unchanged (Fig. 2A-C
Figure 2. Levels of class I HDAC proteins in cortex of CAG140 knock-in mice and HD patients. A) Class I HDAC1 and 3 protein levels in cortices of 8 month-old CAG140 knock-in (KI) and wild-type (wt) mice detected by western. B) Class I HDAC1, HDAC2, and (more ...)
When we assessed the levels of HDAC1 in cortical samples from HD patients and normal individuals (Table 2), HDAC1 levels appeared to be the same (Fig. 2 D, E). However, samples from HD patients and normal subjects were not controlled for disease grade or age. We also observed high variability of HDAC1 levels in human HD samples.
Table 2: Required case information for the human brain tissues analyzed.
Next, we examined levels of HDAC class II proteins at different stages of HD in the mouse models. There were no pronounced differences at the early stage of disease in R6/2 mice (4 weeks) (Fig. 3A, C). However, at end-stage disease in R6/2 mice (12 weeks) we observed statistically significant reductions in levels of class II HDAC proteins, particularly HDAC6 (Fig. 3B, C). Reduced HDAC class II protein levels were not limited to cortex, but were also observed in striata from 9 week-old R6/2 mice (Fig. 3D, E).
Figure 3. Levels of HDAC4, HDAC5, and HDAC6 proteins in cortex and striatum of R6/2 HD mice. A) Western blot showing HDAC4, HDAC5 and HDAC6 protein levels in cortices of 4 week-old R6/2 transgenic (R6/2) and wild-type (wt) mice. B) HDAC4, HDAC5, and HDAC6 (more ...)
In the HD knock-in CAG140 mouse model, we did not observe significant differences in the levels of HDAC 4, 5 and 6 proteins at early (8 month-old) or late (24 month-old) disease stage (Fig. 4). High variability of HDAC5 and HDAC6 protein levels in HD and wild-type aged mice were observed. Similarly, in the human samples, the high variability of HDAC class II proteins did not permit us to detect any clear difference in HD vs. control samples (not shown).
Figure 4. Levels of class II HDAC4, HDAC5, HDAC6 proteins in cortex of CAG140 HD knock-in mice. A) Western blot showing class II HDAC protein levels in cortices of 8 month-old CAG140 knock-in (KI) and wild-type (wt) mice. B) Class II HDAC protein levels (more ...)
To overcome this hurdle we extended our analysis to assess the state of α-tubulin acetylation in HD mouse and human brains. Because α-tubulin is a substrate of the microtubule deacetylase HDAC6, acetylated α-tubulin levels function as a biochemical marker for HDAC6 cellular activity. We examined whether modulation of HDAC6 levels in R6/2 cortices (Fig. 5A, Fig. 6A
) and striata (Fig. 6B
) are translated into higher levels of α-tubulin acetylation. In parallel, we assessed α-tubulin acetylation levels in HD knock-in CAG140 mice (Fig. 5B, Fig. 6D
) and in human samples (Fig. 5C
). We did not observe statistically significant changes in α-tubulin acetylation in any of these tissues (Fig. 5D
). Lastly, we used a pharmacogenomic approach to compare the extent of deacetylase inhibition in cortical extracts from 12 week-old R6/2 and wild-type mice, by using a previously described class II selective inhibitor, MC1568, as a molecular probe ,,,,,,,,
. This compound inhibited deacetylase activities in an identical manner in HD mutant and wild-type cortical samples (Fig. 5E
Figure 5. Evaluation of acetylated a-tubulin levels in HD cortices. A-C) Western analysis of acetylated and total a-tubulin in cortices: A) 12 week-old R6/2 transgenic (R6/2) and wild-type (wt) mice; B) 24 month-old CAG140 knock-in (KI) and wild-type (more ...)
Figure 6. Evaluation of acetylated a-tubulin levels in HD. A) Acetylated and total α-tubulin detected by western in cortices of 4 week-old R6/2 transgenic (R6/2) and wild-type (wt) mice; B) Acetylated and total α-tubulin detected by western (more ...)
Using a similar pharmacological approach, we then tested whether the presence of mutant huntingtin interfered with acetylation/deacetylation of HDAC substrates in a cellular environment physiologically relevant to HD. To that end we tested class-selective HDAC inhibitors in the previously characterized rat embryonic striatal ST14A cells, which express stably integrated mutant (128Q) or wild-type (15Q) 548 amino acid N-terminal fragments of human huntingtin ,
. Compound experiments were conducted in cycling and in differentiated post-mitotic cells. Mutant HD cells demonstrated the expected increase in acetylation in response to treatment with HDAC class I selective MC2070 (A.Mai, S.Valente, M.Tardugno, M.Conte, R.Cirilli, A.Perrone, S.Massa, A.Nebbioso, G.Brosch and L.Altucci, manuscript in preparation) and with class II selective inhibitors MC1568 and MC1575 ,,,,,,,,
(Table 3, 4
Table 3: Chemical structures of tested compounds.
Structures of class II selective HDAC inhibitors MC1568 and MC1575 ,,,,,,,,
and class I selective inhibitor MC2070 (A.Mai, S.Valente, M.Tardugno, M.Conte, R.Cirilli, A.Perrone, S.Massa, A.Nebbioso, G.Brosch and L.Altucci, manuscript in preparation).
Table 4: HDAC inhibition data of tested compounds. Relative HDAC inhibition activities of tested compounds MC1568, MC1575 (class II HDAC selective), and MC2070 (class I HDAC selective), expressed as percent of control (100%).
In details, consistent with the inhibitors class-selectivity, MC2070 and MC1568 respectively increased histone and α-tubulin acetylation in HD treated cells (Fig. 7A). In wild-type cells the effects of the compounds were highly similar (not shown). We next performed side-by-side comparison, in HD mutant and wild-type cells, of the class II selective inhibitor MC1575 effects on α-tubulin acetylation (Fig. 7B). Acetylation levels were determined by western analysis of extracts from cells treated with the compound for 4 hours. Acetylation of α-tubulin increased at the same rate in mutant and wild-type cells, as indicated by quantification of the western blots (Fig. 7C).
Figure 7. Responses of HD cells to HDAC class-selective inhibition with small molecular probes. A) Deacetylase activities inhibition in ST14A HD striatal cells with class I selective MC2070 and class II selective MC1568 inhibitors. Inhibitor effects on (more ...)