Evaluation of 5-(Trifluoromethyl)-1,2,4-oxadiazole-Based Class IIa HDAC Inhibitors for Huntington's Disease

Abstract

Using an iterative structure-activity relationship driven approach, we identified a CNS-penetrant 5-(trifluoromethyl)-1,2,4-oxadiazole (TFMO, 12) with a pharmacokinetic profile suitable for probing class IIa histone deacetylase (HDAC) inhibition in vivo. Given the lack of understanding of endogenous class IIa HDAC substrates, we developed a surrogate readout to measure compound effects in vivo, by exploiting the >100-fold selectivity compound 12 exhibits over class I/IIb HDACs. We achieved adequate brain exposure with compound 12 in mice to estimate a class I/IIb deacetylation EC₅₀, using class I substrate H4K12 acetylation and global acetylation levels as a pharmacodynamic readout. We observed excellent correlation between the compound 12 in vivo pharmacodynamic response and in vitro class I/IIb cellular activity. Applying the same relationship to class IIa HDAC inhibition, we estimated the compound 12 dose required to inhibit class IIa HDAC activity, for use in preclinical models of Huntington's disease.

Figure 1

Examples of HDAC inhibitors containing the TFMO group.

Figure 2

Brain (A) and blood (B) exposure of 12 following oral gavage of compounds to C57/BL6 mice (10, 30, and 100 mg/kg). Left Y axis indicates the measured exposure in the matrix (determined by LC-MS/MS). Right Y axis indicates estimated free exposure, scaled based on f_u values returned in each matrix in vitro. Class IIa cellular IC₅₀ is shown by the red arrowhead; class I/IIb cellular IC₅₀ is shown by the blue arrowhead. Affinity measures for other binding targets identified by the Cerep Diversity panel (K_D range for M1–M4 receptors) indicated by the purple bar, and K_D for HERG channels (IonWorks) indicated by the green arrowhead (SI Table 3).

Figure 3

(A–B) Correlation of HDAC class I/IIb substrate acetylation (broken line) to 12 brain exposure (continuous line). Left axis: concentration of 12 measured by LC-MS/MS in brain (solid line). Right axis: H4K12 normalized to H4 histone levels (broken line) (A) or Pan AcK normalized to tubulin (B) measured via dot blot assay from brain samples. Increasing doses of 12 administered p.o. resulted in increasing “off-target” class I H4K12 acetylation (A) and increasing global acetylation levels (B). (C–D) To confirm specific class I/IIb substrate acetylation, samples from the 2 h time point were probed by Western blot to determine H3, H4, and tubulin acetylation and compared to the 5 min 10 mg/kg samples (C). (D) Quantitation of the pan-AcK band corresponding to H3 (left) and H4 (right) normalized to H3 and H4 total protein band intensity. (E). Quantitation of the pan-AcK band corresponding to tubulin molecular weight normalized to total tubulin protein band intensity (left). Quantitation of tubulin acetylation was equivalent when an Ab specific for acetylated tubulin was used (right). One-way ANOVA with Dunnett’s multiple comparison test: p < 0.05*, p < 0.01**, p < 0.001***, p < 0.0001****.

Figure 4

Oral bid dosing simulation (modeling a dose interval of 10 h), showing simulated total brain concentration of 12. ^a Estimates of final parameters from 10 mg/kg po were used to predict concentrations achieved following 3 mg/kg po. Both doses are predicted to provide brain exposure exceeding class IIa HDAC IC₅₀ for ∼4–6 h post each dose and to have minimal effect on engaging class I/IIb HDACs, with brain exposure failing to reach concentrations > EC₁₀ for class I/IIb substrate acetylation, as determined empirically (SI Figure 3B–C). *The in vivo class IIa IC₅₀ (235 nM) was estimated from the measured cellular IC₅₀ (40 nM) adjusted for 12 F_u of 0.17 in brain homogenate.

Scheme 1. Synthesis of 12

Reagents and conditions: (a) MsCl, Et₃N, DCM, 0 °C to r.t., 18 h. (b) (S)-2-methylpyrrolidine hydrochloride, Cs₂CO₃, DMF, 60 °C, 17 h (35% two steps). (c) HCl, dioxane, DCM, r.t., 2 h (81%). (d) 4-(5-(Trifluoromethyl)-1,2,4-oxadiazol-3-yl)benzoic acid, EDC, HOPO, DIPEA, DCM, r.t., 17 h (97%).