Effects of substrate mechanics on contractility of cardiomyocytes generated from human pluripotent stem cells

Abstract

Human pluripotent stem cell (hPSC-) derived cardiomyocytes have potential applications in drug discovery, toxicity testing, developmental studies, and regenerative medicine. Before these cells can be reliably utilized, characterization of their functionality is required to establish their similarity to native cardiomyocytes. We tracked fluorescent beads embedded in 4.4-99.7 kPa polyacrylamide hydrogels beneath contracting neonatal rat cardiomyocytes and cardiomyocytes generated from hPSCs via growth-factor-induced directed differentiation to measure contractile output in response to changes in substrate mechanics. Contraction stress was determined using traction force microscopy, and morphology was characterized by immunocytochemistry for α-actinin and subsequent image analysis. We found that contraction stress of all types of cardiomyocytes increased with substrate stiffness. This effect was not linked to beating rate or morphology. We demonstrated that hPSC-derived cardiomyocyte contractility responded appropriately to isoprenaline and remained stable in culture over a period of 2 months. This study demonstrates that hPSC-derived cardiomyocytes have appropriate functional responses to substrate stiffness and to a pharmaceutical agent, which motivates their use in further applications such as drug evaluation and cardiac therapies.

Figure 1

Tensile testing data for polyacrylamide hydrogels shows a linear relationship between elastic modulus and crosslinker concentration. The concentration of acrylamide monomer was held constant at 10% and the concentration of bisacrylamide crosslinker was varied between 0.03–0.6%. Substrate stiffness increased linearly with bisacrylamide concentration over the tested range of compositions. Linear regression demonstrated a significantly nonzero slope (m = 159.4 ± 4.6, P < 0.0001). n = 6–13 for each hydrogel composition and error bars represent SEM.

Figure 2

Methods and sample data for obtaining contraction stress measurements. (a) Flow cytometry data shows a typical H9-derived cardiomyocyte population of >90% purity on day 15 of differentiation, demonstrated by the percentage of cells expressing cardiac Troponin T (cTnT). (b) Schematic of polyacrylamide hydrogel cross-section after surface treatment and cell seeding. (c) Merged image shows a contracting D30 19-9-11-derived cardiomyocyte and green fluorescent beads embedded in the 76.0 kPa substrate beneath the cell. The cell was at its maximum point in the contraction cycle. Scale bar = 20 μm. (d) Contraction stress map for the cell pictured in (c) shows the range and localizations of contraction stresses.

Figure 3

Contraction stress of cardiomyocytes increased with substrate stiffness; bead displacement and beating rate did not increase with stiffness. (a) Average contraction stress of neonatal rat, D30 H9-derived, and D30 19-9-11-derived cardiomyocytes increased with substrate stiffness. Each data point represents an average of the absolute values of contraction stresses over area for a single cell at the maximum point of its contraction cycle. Mean contraction stress values were significantly affected by stiffness via one-way ANOVA (overall P = 0.0018 for rat and overall P < 0.0001 for D30 H9 and D30 19-9-11). ***(P < 0.001) and *(P < 0.05) indicate statistically significant differences relative to neonatal rat cardiomyocytes on the same stiffness. (b) Maximum contraction stress of neonatal rat, D30 H9-derived, and D30 19-9-11-derived cardiomyocytes increased with substrate stiffness. Each data point represents the upper limit of the range of contraction stresses generated by a single cell at the maximum point of its contraction cycle. ***(P < 0.0001) and ^†(P < 0.05) indicate statistically significant differences relative to neonatal rat and D30 19-9-11-derived cardiomyocytes, respectively, on the same stiffness. (c) Displacement of fluorescent beads occurred to a similar extent on all stiffnesses. Each data point represents an average of the absolute values of bead displacement for a single cell, including both near- and far-field beads. For all cell lines, substrate stiffness significantly affected bead displacement using one-way ANOVA (overall P < 0.0001, P = 0.0001, P = 0.0003 for rat, D30 H9, and D30 19-9-11 resp.), with an overall significant decreasing (rat and D30 H9) or nonsignificant increasing (D30 19-9-11) trend via linear regression. (d) Beating rate was similar on all stiffnesses. Substrate stiffness did not significantly affect beating rate via one-way ANOVA for each cell line (overall P = 0.92, P = 0.18, P = 0.07 for rat, D30 H9, and D30 19-9-11, resp.). For (a)–(d), n = 7–11 cells for each stiffness and error bars represent SEM.

Figure 4

Morphological characterization of D30 19-9-11-derived cardiomyocytes on polyacrylamide hydrogels. Cells were seeded onto the hydrogels at 1 month postdifferentiation, fixed 24 hours later, and immunostained for α-actinin. Morphology was characterized using CellProfiler software. (a) Substrate stiffness significantly affected cell area (overall P < 0.0001 via one-way ANOVA). Cell area peaked on the 49.4 kPa hydrogel. ***(P < 0.001) and *(P < 0.05) indicate statistically significant differences. (b) Substrate stiffness did not significantly affect eccentricity (overall P = 0.12 via one-way ANOVA). For (a) and (b), n = 81–112 cells per stiffness and error bars represent SEM. (c) Representative images show sarcomere organization on all stiffnesses. α-actinin is shown in red, and nuclei (stained with Hoechst) are shown in blue. Scale bar = 10 μm.

Figure 5

Isoprenaline treatment increased beating rate but not contraction stress of D30 H9-derived cardiomyocytes. Cells were seeded onto polyacrylamide hydrogels at 1 month postdifferentiation and imaged 24 hours later to obtain contraction stress and beating rate. One population was untreated, and the other was treated with 9 μM isoprenaline for 5 minutes prior to imaging. (a) Average contraction stress was not significantly affected by isoprenaline treatment (overall P = 0.40 for effect of treatment via two-way ANOVA). n = 8-9 cells per stiffness for untreated and 5–7 cells per stiffness for isoprenaline treated. (b) Beating rate significantly increased upon isoprenaline treatment (overall P < 0.0001 for effect of treatment via two-way ANOVA). **(P < 0.01) and *(P < 0.05) indicate statistically significant differences. n = 9–11 cells per stiffness for untreated and 6–9 cells per stiffness for isoprenaline treated. For (a) and (b), error bars represent SEM.

Figure 6

Contraction stress remained stable with time past differentiation, while beating rate changed on soft substrates. H9-derived cardiomyocytes were seeded onto polyacrylamide hydrogels at 30 and 60 days postdifferentiation and imaged 24 hours later to obtain contraction stress and beating rate. (a) Average contraction stress increased with substrate stiffness but was not significantly different on any stiffness at D30 and D60 (overall P = 0.29 for effect of time via two-way ANOVA). (b) Maximum contraction stress increased with substrate stiffness but was not significantly different on any stiffness at D30 and D60 (overall P = 0.14 for effect of time via two-way ANOVA). For (a) and (b), n = 8-9 cells per stiffness for D30 and 9–13 cells per stiffness for D60. (c) Stiffness did not significantly affect beating rates of D60 H9-derived cardiomyocytes (overall P = 0.13 via one-way ANOVA), but beating rate increased on soft substrates between D30 and D60. *(P < 0.05) indicates statistically significant differences. n = 8–11 cells per stiffness for D30 and 10–13 cells per stiffness for D60. For (a)–(c), error bars represent SEM.