Determining properties of human‐induced pluripotent stem cell‐derived cardiomyocytes using spatially resolved electromechanical metrics

Figures

• 

  Figure 1. Open‐well cardiac microtissues (Samples C1–C6) (A) TPE trough design, (B) TPE cut‐outs (approx. 2.5 × 3.5 mm2) placed in a 96‐well plate and (C) schematic drawing (not to scale) of a TPE sample in a 96‐well plate (1). For cell loading, a drop of cell suspension is added covering the troughs (2). After cell packing into the troughs, fresh culture media are added (3). (D) CM1 cell loading (Day 0, left) and tissue formation (Day 8, right).

• 

  Figure 2. Illustrations of the considered biomarkers. The beat rate (BR), the action potential duration (APD) and the conduction velocity (CV) are computed from recordings of voltage sensitive dyes, and the displacement amplitude (DA) is computed from brightfield images as described in eqn (1).

• 

  Figure 3. Illustration of components of the applied single‐cell model. The left panel illustrates the transmembrane currents and the intracellular calcium fluxes, in addition to the five considered intracellular compartments. The right panel illustrates the Markov model used to represent the myosin–actin cross‐bridge dynamics. This Markov model has four states, the non‐permissive resting state (R), the permissive detached state (D), the cross‐bridge pre power stroke (X) and the cross‐bridge post power stoke state (P).

• 

  Figure 4. Illustrations of the cell geometry and possible cell connections in the spatial cell collection model The centre point of cell i is denoted by ϕi=p4i and the possible connection points are denoted by pqi, for q=0,1,2,3,5,6,7,8.

• 

  Figure 5. Illustration of the seven chips (i.e. cell collections) C1–C7 considered in this study. The left panel shows the average (over time) of the optical measurements using voltage sensitive dyes for the baseline case with no drug present. The right panel shows the geometry used in the computational model for each of the different cell collections.

• 

  Figure 6. More detailed illustration of the model cell distribution in Chip C2. The brown round objects are pillars, and the orange rectangles are hiPSC‐CMs.

• 

  Figure 7. Illustration of a slow electrical wave travelling through Chip C1. The left panel shows optical measurements from an experiment, and the right panel shows the model solution. The lower first and third panels show the average transmembrane potential in seven local areas of the chip. The upper panel illustrates the local areas used to define the transmembrane potential in the lower panel. The second and fourth lower panels compare the action potential in the fourth local area (grey) to the average trace over all seven areas (orange).

• 

  Figure 8. Illustration of (almost) synchronous cells in Chip C4. In the experiment (left panel), the action potential upstroke taken from four locations in the chip (see upper panel) happens almost synchronously (CV > 10 cm/s). The global average action potential or global average full voltage trace (taken over the entire chip) is almost identical to local averages (taken in the second local area). In the model, we assume that the cells beat synchronously. We compare the global average, a local average and the solution of a single‐cell simulation. All three solutions are virtually indistinguishable.

• 

  Figure 9. Measured and simulated voltage and displacement traces for the cells in Chip C4 in the baseline case and after the application of 0.1 µM of the L‐type calcium channel blocker, nifedipine. In the simulations, the effect of the drug is represented by reducing gCaL by 50%. Furthermore, in the single‐cell simulation, the displacement is defined as the shortening of the cell length, whereas for the experiment and the spatial simulation, the displacement is defined as described in the ‘Computing displacement traces from brightfield images’ section.

• 

  Figure 10. Sensitivity of the biomarkers to changes in the biophysical parameters gKr, gNa, gCaL, kRD, kDX, kXD, kPD and Gi,j. The figures display the change of biomarker value (in percentage) resulting from a 25% reduction in the parameter values. The biomarkers are computed from model simulations using the parameterization and spatial cell distribution of Chip C1 (see Fig. 7). In the upper panel, we consider non‐synchronous cells with an excitation wave initiated by a 10% weaker gK1 in the leftmost cells. In the lower panel, we consider synchronously beating cells with uniform cell properties.

• 

  Figure 11. Flecainide. Average action potential upstroke in four local areas of Chip C2 in the baseline case and after application of 1 and 10 µM of flecainide. The upper panel illustrates the local areas used to define the action potential upstrokes in the lower panel. The left panels show optical measurements from an experiment, and the right panels show the model solution. The effect of the drug is modelled as described in Table 9.

• 

  Figure 12. Flecainide. Snapshots of the model solutions showing electrical waves traversing Chip C2 after the application of 1 and 10 µM of flecainide. In the baseline case, no waves are present because the cells in the chip beat synchronously.