Human Engineered Cardiac Tissues Created Using Induced Pluripotent Stem Cells Reveal Functional Characteristics of BRAF-Mediated Hypertrophic Cardiomyopathy

Abstract

Hypertrophic cardiomyopathy (HCM) is a leading cause of sudden cardiac death that often goes undetected in the general population. HCM is also prevalent in patients with cardio-facio-cutaneous syndrome (CFCS), which is a genetic disorder characterized by aberrant signaling in the RAS/MAPK signaling cascade. Understanding the mechanisms of HCM development in such RASopathies may lead to novel therapeutic strategies, but relevant experimental models of the human condition are lacking. Therefore, the objective of this study was to develop the first 3D human engineered cardiac tissue (hECT) model of HCM. The hECTs were created using human cardiomyocytes obtained by directed differentiation of induced pluripotent stem cells derived from a patient with CFCS due to an activating BRAF mutation. The mutant myocytes were directly conjugated at a 3:1 ratio with a stromal cell population to create a tissue of defined composition. Compared to healthy patient control hECTs, BRAF-hECTs displayed a hypertrophic phenotype by culture day 6, with significantly increased tissue size, twitch force, and atrial natriuretic peptide (ANP) gene expression. Twitch characteristics reflected increased contraction and relaxation rates and shorter twitch duration in BRAF-hECTs, which also had a significantly higher maximum capture rate and lower excitation threshold during electrical pacing, consistent with a more arrhythmogenic substrate. By culture day 11, twitch force was no longer different between BRAF and wild-type hECTs, revealing a temporal aspect of disease modeling with tissue engineering. Principal component analysis identified diastolic force as a key factor that changed from day 6 to day 11, supported by a higher passive stiffness in day 11 BRAF-hECTs. In summary, human engineered cardiac tissues created from BRAF mutant cells recapitulated, for the first time, key aspects of the HCM phenotype, offering a new in vitro model for studying intrinsic mechanisms and screening new therapeutic approaches for this lethal form of heart disease.

Fig 1. Characterization of wild-type and BRAF iPSC-hECTs.

(A) Sorting paradigm for construction of defined iPSC-hECTs; (B) Photograph of wild-type (left) and mutant (right) tissues after 6 days in culture; (C) longitudinal section of wild-type (left) and mutant (right) tissues stained with hematoxylin and eosin after 12 days in culture; (D) Post deflections of both tissue types on day 6 at 2 Hz pacing frequency; (E) Cross-sectional area (mean±SD) of wild-type (open bars, n = 7) and mutant (filled bars, n = 4) tissues; (F) Molecular analysis of mutant and wild-type tissues performed on day 12 (n = 3 for each tissue type).

Fig 2. Twitch force characteristics of wild-type and BRAF-mutant hECTs.

(A) Cartoon of iPSC-derived hECT force data illustrating twitch characteristic parameters. (B-G) Mean (±SD) twitch parameters for wild-type (open bars, n = 7) and BRAF mutant (solid bars, n = 5) iPSC-hECTs tested on culture days 6 and 11, including developed force (B), 50% twitch duration (C), time to 50% contraction (D), time to 50% relaxation (E) and the maximum rates of contraction (F) and relaxation (G). * p < 0.05, ** p < 0.01, *** p < 0.001.

Fig 3. Electrical properties of of wild-type and BRAF-mutant hECTs.

(A) Developed force versus frequency relationship of iPSC-derived hECTs at culture day 6 (left) and day 11 (right); (B) maximum pacing frequency of wild-type and mutant tissues; (C) spontaneous beating frequency; (D) The minimum voltage necessary for pacing for both tissues; (E) Relative chronotropic response to isoproterenol (normalized to baseline beating frequency) with fitted nonlinear regression model. (F) Twitch rate variability plots of each tissue type with a paced control for comparison. Each color represents a different tissue. (G) Quantification of twitch rate variability by the mean distance of each point on the twitch rate variability plots from each cluster centroid. Error bars represent standard error (A) or standard deviation (B-G) Open symbols and bars are wild-type hECTS (n = 7), filled symbols and bars are BRAF-mutant hECTs (n = 5). * p < 0.05, ** p < 0.01, *** p < 0.001.

Fig 4. Investigating longitudinal changes in hECT functional phenotype.

(A) Principal component analysis and k-means clustering of tissues paced at 2 Hz on day 6 (left) and day 11 (right). (B) Developed force of the wild-type tissues increased significantly over day 6 and day 11, while mutant tissues did not. The lack of a significant change from day 6 to day 11 appeared to be due to a greater increase in diastolic force of mutants from day 6 to 11 than wild-type tissues, while systolic force increased approximately by the same amount for both tissue types. (C) Passive Young’s modulus determined by uniaxial stretch measurements on days 12–15 (n = 4 per tissue type). * p < 0.05 between tissues types, † p < 0.05 between day 6 and day 11 for wild-type tissues.