Aligned human cardiac syncytium for in vitro analysis of electrical, structural, and mechanical readouts

Abstract

Human pluripotent stem cell-derived cardiomyocytes (hPSC-CMs) have emerged as an exciting new tool for cardiac research and can serve as a preclinical platform for drug development and disease modeling studies. However, these aspirations are limited by current culture methods in which hPSC-CMs resemble fetal human cardiomyocytes in terms of structure and function. Herein we provide a novel in vitro platform that includes patterned extracellular matrix with physiological substrate stiffness and is amenable to both mechanical and electrical analysis. Micropatterned lanes promote the cellular and myofibril alignment of hPSC-CMs while the addition of micropatterned bridges enable formation of a functional cardiac syncytium that beats synchronously over a large two-dimensional area. We investigated the electrophysiological properties of the patterned cardiac constructs and showed they have anisotropic electrical impulse propagation, as occurs in the native myocardium, with speeds 2x faster in the primary direction of the pattern as compared to the transverse direction. Lastly, we interrogated the mechanical function of the pattern constructs and demonstrated the utility of this platform in recording the strength of cardiomyocyte contractions. This biomimetic platform with electrical and mechanical readout capabilities will enable the study of cardiac disease and the influence of pharmaceuticals and toxins on cardiomyocyte function. The platform also holds potential for high throughput evaluation of drug safety and efficacy, thus furthering our understanding of cardiovascular disease and increasing the translational use of hPSC-CMs.

Keywords: anisotropic conduction; human pluripotent stem cell-derived cardiomyocytes; microcontact printing; substrate stiffness.

Figure 1:

(A) Schematic of the experimental timeline. Red numbers denote days after hiPSC-CM differentiation while black numbers represent experimental time points. (B) Lactate purified hiPSC-CMs patterned on Matrigel lanes of varying widths on 10kPa PDMS and stained with α-actinin (green). (C) Bright field image of hiPSC-CMs patterned on 30 μm lanes spaced apart by 30 μm. Red arrows indicate merged regions between adjacent lanes. (D) hiPSC-CMs cultured on the 90° mesh pattern for 12 days with red stars denoting areas of pattern deterioration. (E) hiPSC-CMs cultured on the 15° chevron pattern for 12 days and no evidence of gaps in the pattern. (F) Quantification of the amount of gaps present in the 90° mesh and 15° chevron pattern per the available area of micropatterned ECM in the bright field image. *p < .001, **p < 0.05, two-way ANOVA with post hoc Tukey tests. Scale bars = 20 μm in B, 100 μm for C-E. N = 9 each for 90° mesh and 15° chevron pattern in Panel F.

Figure 2:

(A) Schematic of the experimental timeline. Red numbers represent days after hiPSC-CM differentiation while black numbers indicate the start and end of culture. hiPSC-CMs cultured for 18 days on the (B) monolayer control and (D) 15°chevron pattern and stained for Green = α-actinin, White = N-cadherin, Blue = DAPI. Scale bars = 50 μm. White dashed boxes in B and D represent SGFT myofibril alignment histogram output for (C) monolayer control and (E) 15° chevron pattern. (F) Myofibril alignment and (G) hiPSC-CM aspect ratio for patterned construct and monolayer control. (F) Percent of myofibrils aligned within 10 degrees of the primary axis. *p < .05, unpaired t-test. Myofibril alignment n = 6, cardiomyocyte aspect ratio n = 4.

Figure 3:

Optical mapping of the (A) human left ventricle (modified with permission from (Glukhov et al., 2012)) and (B) monolayer and patterned hiPSC-CM constructs. Field of view for optical mapping results is 1cm × 1cm. Conduction velocities for monolayer and patterned constructs when paced at 1Hz. (D) Anisotropic conduction determined by comparison of the longitudinal and transverse conduction velocity speeds. *p < .05, **p < .001 unpaired t-test. Monolayer n = 14, pattern n = 10.

Figure 4:

Workflow of contractile strain analysis. (A) From bright field videos of a hiPSC-CM contraction displacements and strains are calculated through the image, examples of which are shown here. A mask, derived from the bright field image, is used to exclude data from areas not occupied by cells. (B) The second principal strain was calculated for each location containing cells and then averaged for each frame of the video and the peak value over time, identified as the maximum contractile strain, was used as a comparative measure of the strain for that sample. (C) 2^nd principal strain was used to compute contractile strain for monolayer and patterned hiPSC-CMs. No statistical difference between monolayer or patterned hiPSC-CMs for the three time points tested. Monolayer n = 15, pattern n = 19 per time point.