Leak current, even with gigaohm seals, can cause misinterpretation of stem cell-derived cardiomyocyte action potential recordings

Abstract

Aims: Human-induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) have become an essential tool to study arrhythmia mechanisms. Much of the foundational work on these cells, as well as the computational models built from the resultant data, has overlooked the contribution of seal-leak current on the immature and heterogeneous phenotype that has come to define these cells. The aim of this study is to understand the effect of seal-leak current on recordings of action potential (AP) morphology.

Methods and results: Action potentials were recorded in human iPSC-CMs using patch clamp and simulated using previously published mathematical models. Our in silico and in vitro studies demonstrate how seal-leak current depolarizes APs, substantially affecting their morphology, even with seal resistances (Rseal) above 1 GΩ. We show that compensation of this leak current is difficult due to challenges with obtaining accurate measures of Rseal during an experiment. Using simulation, we show that Rseal measures (i) change during an experiment, invalidating the use of pre-rupture values, and (ii) are polluted by the presence of transmembrane currents at every voltage. Finally, we posit that the background sodium current in baseline iPSC-CM models imitates the effects of seal-leak current and is increased to a level that masks the effects of seal-leak current on iPSC-CMs.

Conclusion: Based on these findings, we make recommendations to improve iPSC-CM AP data acquisition, interpretation, and model-building. Taking these recommendations into account will improve our understanding of iPSC-CM physiology and the descriptive ability of models built from such data.

Keywords: Arrhythmias; Computer simulation; Induced pluripotent stem cells; Ion channels; Patch clamp.

Graphical abstract

Figure 1

R _seal cannot be measured directly once access is gained. Once access is gained, we can only measure the combined resistance R_in, which is equal to the parallel resistances of R_seal and R_m [Eq. (4)]. The presence of R_m introduces uncertainty when R_in is used to approximate R_seal, making it difficult to accurately correct for leak current effects. For simplicity, we have omitted other elements of this patch clamp diagram (e.g. series resistance and capacitance).

Figure 2

Effect of R_seal on Kernik and Paci APs. Simulations from the Kernik + leak (A) and Paci + leak (B) models, each with capacitance set to 50 pF (the experimental average), and R_seal set to values from 1 to 10 GΩ. The dashed (red) trace shows a baseline (leak-free) simulation. Four AP morphology metrics for the Kernik (C) and Paci (D) models are plotted against R_seal (displayed on log-scaled x-axis): MP, dV/dt_max, APD₉₀, and CL. Grey boxes denote the R_seal values where the Kernik model is non-spontaneous. Abbreviations: APD₉₀, action potential duration at 90% repolarization; CL, cycle length; dV/dt_max, maximum upstroke velocity.

Figure 3

Effect of R_seal on ToR-ORd adult cardiomyocyte APs at 50 and 153 pF. Simulations from the ToR-ORd + leak model paced at 1 Hz with C_m set to 50 (A) and 153 pF (B), and R_seal set to values from 1 to 10 GΩ. The dashed (red) trace shows a baseline (leak-free) simulation. Three AP morphology metrics for the 50 and 153 pF models are plotted against R_seal (displayed on log-scaled x-axis): APD₉₀, action potential duration at 90% repolarization; dV/dt_max, maximum upstroke velocity; MP, minimum potential.

Figure 4

R _in changes during iPSC-CM experiments. (A) Distribution of initial R_in measurements from iPSC-CMs acquired with a +5 mV step from 0 mV. (B) The percentage change in R_in plotted against the time elapsed between R_in measurements. The interval between measurements ranged from 1 to 10 min. Time was recorded to the nearest minute, leading to the appearance of banding in the ΔTime measure.

Figure 5

Ignoring the presence of I_f makes it impossible to accurately measure R_seal after gaining access. (A) Voltage clamp data acquired from an iPSC-CM before and after treatment with quinine, which is expected to block 32% of I_f at the concentration used. (B) Kernik model response at baseline and with 32% block of I_f. (C) Kernik + leak voltage clamp simulations conducted with R_seal = 1 GΩ, g_K1 reduced by 90%, and g_f set to 0 (solid line), 0.0435 (dotted line), or 0.087 nS/pF (dashed line). A voltage step from −80 to −75 mV was applied, as is commonly used to estimate R_in. This R_in value is sometimes used to approximate R_seal when the holding potential is near −80 mV. The amplifier-measured (I_out), total transmembrane (I_ion), and leak currents (I_leak) are displayed. The R_in values calculated based on ΔI_out are 2.03, 1.50, and 1.16 GΩ for the 0, 0.0435, and 0.087 nS/pF simulations, respectively. (D) R_in values are plotted against holding potential for Kernik + leak models with R_seal = 1 GΩ and g_f equal to 0, 0.0435, or 0.087 nS/pF. The horizontal dotted line shows the true simulated R_seal value of 1 GΩ.

Figure 6

R _in predictions of R_seal are overestimated at the reversal potential for leak current. (A) The current response (I_out) for Kernik + leak models with a 1 GΩ seal and g_f of 0 (solid line), 0.0435 (dotted line), or 0.087 nS/pF (dashed line) to a 50 ms + 5 mV voltage clamp step from 0 mV (top) or −80 mV (bottom). (B) Effect of R_seal on R_in measures for models with g_f set to 0 (solid), 0.0435 (dotted), or 0.087 nS/pF (dashed). R_in was calculated with Eq. (3). The +5 mV voltage steps were taken from either 0 or −80 mV. The R_seal = R_in line (dotted, without symbols) is provided as a reference for when R_in correctly predicts R_seal. The 0 mV lines are overlapping, illustrating that R_in is not sensitive to g_f at this voltage. The g_f = 0.0875 nS/pF model at −80 mV provides the best estimate of R_seal.

Figure 7

Cells appeared phenotypically heterogeneous, with uncorrelated variation in g_in and C_m. (A) Current clamp recordings from three cells show phenotypic heterogeneity: non-spontaneous (green), spontaneous AP with short APD (teal), and spontaneous AP with long APD (red). (B) MP and APD₉₀ for spontaneously beating cells (n = 25). Note the broken x-axis that allows us to display an outlying data point. (C) The relationship between C_m and g_in for all cells (n = 37). Non-spontaneous cell data points are denoted with squares, whilst spontaneous are circles. APD₉₀, action potential duration at 90% repolarization; MP, minimum potential.

Figure 8

Relationship between g_in/C_m and AP biomarkers. (A) g_in/C_m plotted against MP. Spontaneously firing cells are denoted as teal circles and non-firing cells as red squares. Linear regression fits to data from spontaneous (teal dashed, R = 0.47, P < 0.05), and non-firing (red dotted, R = 0.76, P < 0.05) cells are overlaid on the plot. No statistically significant relationship was found between g_in/C_m and APD₉₀ (B), CL (C), or dV/dt_max (D). APD₉₀, action potential duration at 90% repolarization; CL, cycle length; dV/dt_max, maximum upstroke velocity; MP, minimum potential.

Figure 9

A simulated example of how leak can be absorbed into background currents: Kernik baseline model fit to Kernik + leak model. The I_bNa and I_bCa conductances (g_bNa and g_bCa) of the baseline Kernik model were fit to a Kernik + leak model (i.e. original + leak) with R_seal set to 5 GΩ using a genetic algorithm. (A) The conductances for all individuals (teal circles) and the best fit individual (red square) from the last generation. (B) Traces from the original baseline Kernik + leak model with a 5 GΩ seal (teal solid), the best fit model from the last generation (red dashed), and the original baseline Kernik model (green dotted).