Discovery of small molecule modulators of the thin filament that alter the contractile profile in a hypertrophic cardiomyopathy engineered tissue model

Figures

• 

  Figure 1. Identification of a thienopyrimidine (TP) series of desensitizers.A, HTS screening funnel using ASMS followed by a series of functional biochemical assays. B, Compound 1, a representative thienopyrimidine (TP) series desensitizer. The reported compound EC50 was measured by holding the free Ca2+ concentration at pCa50 (2.1 μM) and normalizing to DMSO control wells. C, skinned left ventricular papillary C57BL/6N mouse muscle fibers were treated with DMSO or 30 μM Compound 1 for 30 min to assess the effect on the force-pCa relationship (N = 3). Data is mean ± SD. TP, thienopyrimidine.

• 

  Figure 2. Optimization of Pharmacological Properties of the Thienopyrimidine (TP) Series. The table illustrates for six compounds how modifications at a core sulfur or oxygen (X), R2, or R4 vector alter myofibril potency, pharmacokinetic properties, and selectivity profile against phosphodiesterase PDE3, PDE4, PDE5 and the human ether-à-go-go-related (hERG) ion channel. Myofibril rat and non-human primate (NHP), hERG, RRCK, hHEP, and LogD values are reported as the arithmetic mean ± SD of at least three independent experiments unless otherwise indicated in parenthesis. PDE 3A1, 4D3, and 5A1 (PDE3/4/5) phosphodiesterase activity assays are reported as fold selectivity over myofibril EC50. hERG, human ether-à-go-go-related; NHP, non-human primate.

• 

  Figure 3. Isothermal titration calorimetry (ITC) of reconstituted bovine cardiac thin filament complex (TFC), actin-tropomyosin complex, and troponin complex with Compound 4.A, representative isotherms for TFC (left), actin-tropomyosin (middle), and troponin complex (right). Significant binding is only observed with the full TFC, shown with fitted thermodynamic parameters of binding using a one-site model. The mean KD obtained from five independent experiments for the TFC was 247 ± 198 nM (B) Mean thermodynamic properties (ΔH, −TΔS, and ΔG) of Compound 4 binding to TFC were −11.4 ± −3.7 kcal/mol (ΔH), 1.98 ± 3.23 kcal/mol (−TΔS), and −9.23 ± 0.63 kcal/mol (ΔG). All titrations were carried out in 40 mM MOPS pH 7.0, 70 mM NaCl, 10 mM MgCl2, 2 mM ADP, 2.1 μM CaCl2, and 1% DMSO. ITC, isothermal titration calorimetry; TFC, thin filament complex.

• 

  Figure 4. Compound 3, 5, and 6 have distinct force-pCa pharmacology profiles in mouse skinned papillary fibers. Each fiber was run twice in sequential order from low (pCa10) to high (pCa 4.5) calcium containing assay buffer. The first run was in the presence of DMSO (Run #1), and the second run was either DMSO again or compound at a set concentration (Run #2). %Max Tension is the tension at the highest calcium concentration (pCa 4.5) normalized to the initial DMSO run (Run #1) of that fiber. ΔpCa50 was calculated as pCa50 of Run #2 - pCa50 of Run #1 (DMSO), where pCa50 is the calcium concentration that elicits 50% activation of the fiber. Three fibers were run for each dose of compound and six fibers for DMSO (0 μM). A, effect of indicated compound treatment on %Max Tension. Data are shown on graphs with mean ± SD indicated and were analyzed by One Way ANOVA and Tukey HSD test to compare %Max Tension between the compound groups. B, graphed data is comparing the ΔpCa50 and %Max Tension @ pCa 4.5 calculated for each individual compound- or DMSO-treated fiber. For a given Compound, ΔpCa50 versus % Max Tension @ pCa 4.5 were analyzed with a linear model, and contrasts were performed to compare…

• 

  Figure 5. Compound 3 and 5 alter isometric twitch contractile properties in an engineered heart tissue (EHT) model of HCM. Compound 3 and 5 were compared in dose-response against EHT seeded with human induced pluripotent stem cell (hiPSC)-derived cardiomyocytes heterozygous for the HCM-associated myosin mutation MYH7 R403Q+/−. A, peak isometric contractile force, (B) time from stimulus to peak isometric force (TTP), and time from peak isometric force to (C) 50% relaxation (RT50) were collected for N = 8 EHT per compound. Data for A-C are all shown as scatter plots with mean ± SD indicated and were analyzed by linear mixed effects model with an AR(1) covariance structure and a random effect for each fiber. Contrasts were performed to compare between the doses. ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001. D, comparison of the relationship between baseline normalized force to baseline normalized RT50 for Compound 3 (black) and Compound 5 (blue). Data were analyzed by linear mixed effects model with an AR(1) covariance structure and a random effect for each fiber to compare the RT50 ∼ Force slope while controlling for compound and dose. Contrasts were fit to compare between the slopes for each…