NanoMEA: A Tool for High-Throughput, Electrophysiological Phenotyping of Patterned Excitable Cells

Abstract

Matrix nanotopographical cues are known to regulate the structure and function of somatic cells derived from human pluripotent stem cell (hPSC) sources. High-throughput electrophysiological analysis of excitable cells derived from hPSCs is possible via multielectrode arrays (MEAs) but conventional MEA platforms use flat substrates and do not reproduce physiologically relevant tissue-specific architecture. To address this issue, we developed a high-throughput nanotopographically patterned multielectrode array (nanoMEA) by integrating conductive, ion-permeable, nanotopographic patterns with 48-well MEA plates, and investigated the effect of substrate-mediated cytoskeletal organization on hPSC-derived cardiomyocyte and neuronal function at scale. Using our nanoMEA platform, we found patterned hPSC-derived cardiac monolayers exhibit both enhanced structural organization and greater sensitivity to treatment with calcium blocking or conduction inhibiting compounds when subjected to high-throughput dose-response studies. Similarly, hPSC-derived neurons grown on nanoMEA substrates exhibit faster migration and neurite outgrowth speeds, greater colocalization of pre- and postsynaptic markers, and enhanced cell-cell communication only revealed through examination of data sets derived from multiple technical replicates. The presented data highlight the nanoMEA as a new tool to facilitate high-throughput, electrophysiological analysis of ordered cardiac and neuronal monolayers, which can have important implications for preclinical analysis of excitable cell function.

Keywords: Multielectrode arrays; cardiomyocyte; electrophysiology; iPSC; nanotopography; neuron.

Figure 1:. Design, fabrication, and characterization of the multiwell, nanotopographically-patterned MEA devices.

(a) High-throughput nanoMEA concept. Each well of the multiwell plate (i) supports independent cell cultures for high-throughput analysis. Within each well (ii), the electrode bed facilitates recording of field potentials generated by the overlying cells. Nanotopography (iii) is applied to each well (or a subset of wells) to promote cellular alignment and functional development. Captured signals are relayed, via an amplifier (iv), to a software program for subsequent analysis (v). (b) Low magnification image of multiwell MEA plate with nanotopography applied to each well. (c) Nafion nanotopography applied to a single well of a 48-well MEA plate. The presence of the nanoscale features causes light diffraction on the surface, giving the patterns a green-orange color in this image (white arrow). (d) SEM image of Nafion nanotopography. (e) NanoMEA fabrication schematic. Pristine wells are first treated with PEDOT to improve the sensitivity of the base electrode. A drop of Nafion resin is then applied to the substrate, and a PDMS mold is pressed into it. After overnight curing, the PDMS mold is removed to reveal Nafion topographic substrates underneath. (f) Percentage of MEA electrodes from which hPSC-CM signal detection could be clearly distinguished above background noise (n = 3 wells per condition, 64 electrodes per well). (g) Quantification of noise recorded from bare, flat Nafion, and patterned Nafion electrodes with and without PEDOT treatment (n = 32 independently analyzed electrodes per condition). Cells maintained on untreated and flat Nafion coated MEAs for 7 days were assessed for differences in: (h) depolarizing spike amplitude, (i) depolarizing spike slope, and (j) field potential duration corrected for beat period (FPDc) (n = 3 wells per condition). *p < 0.05.

Figure 2:. Structural and functional properties of patterned human cardiomyocytes on nanoMEAs.

(a) Immunostained image of hPSC-CMs on flat Nafion substrates. Inset shows detail of the sarcomeric structures present. (b) Immunostained image of hPSC-CMs on nanotopographically-patterned Nafion substrates. Inset shows detail of the sarcomeric structures present. (c) Measurement of conduction velocity (CV) across hPSC-CM monolayers on flat and nanotopographically-patterned MEAs (n = 21 wells for flat and 14 for patterned). Both transverse conduction velocity (TCV) and longitudinal conduction velocity (LCV) were measured and recorded. (d) Quantification of pixel intensity in images collected from hPSC-CM cultures stained with a primary antibody that targets Cx43 (n = 30 images for flat and 33 for patterned, collected from 2 independent cultures). (e) Representative traces (averaged across 10 beats) recorded from hPSC-CM monolayers on flat MEAs and subjected to increasing doses of bepridil. (f) Representative traces recorded from hPSC-CM monolayers on nanoMEAs and subjected to increasing doses of bepridil. (g) Normalized dose response curve illustrating effect of increasing concentrations of bepridil on the FPDc of unpatterned and patterned hPSC-CMs. The R² values for the unpatterned and patterned cultures were 0.58 and 0.55, respectively (n = 10 wells per dose per condition). (h) Normalized dose response curve illustrating the effect of increasing concentrations of cisapride on the FPDc of unpatterned and patterned hPSC-CMs. The R² values for the unpatterned and patterned cultures were 0.46 and 0.51, respectively (n = 8 wells per dose per condition). (i) Average change in fluorescence intensity recorded from Fluo-4-treated hPSC-CMs maintained on patterned surfaces with and without 1 μM bepridil exposure (n = 11 (untreated) and 16 (treated)). (j) Time between calcium transient peak fluorescence and 90% return to baseline in Fluo-4-treated, flat and patterned hPSC-CMs with and without 1 μM bepridil exposure (n = 11 (untreated) and 16 (treated)). (k) Normalized dose response curves illustrating the effect of increasing concentrations of carbenoxolone on the conduction velocity of unpatterned hPSC-CM monolayers. Dose response curves were calculated from analysis of propagation speeds in both the transverse (TCV) and longitudinal (LCV) directions. The R2 values for TCV and LCV curve fits were 0.23 and 0.11, respectively (n = 4 wells per dose per condition). (l) Normalized dose response curves illustrating the effect of increasing concentrations of carbenoxolone on the TCV and LCV of patterned hPSC-CM monolayers. The R2 values for TCV and LCV curve fits were 0.21 and 0.28, respectively (n = 4 wells per dose per condition). In all experiments, underlying nanotopography was organized to run longitudinally. *p < 0.05, **p < 0.005, and ***p < 0.001.

Figure 3:. Structural and functional properties of patterned human neurons on nanoMEAs.

(a) Bright-field image of low-density hPSC-Ns maintained on flat Nafion substrates. (b) Bright-field image of low-density hPSC-Ns maintained on nanotopographically-patterned Nafion substrates. (c) Measurement of the number of spikes recorded per network burst from high-density hPSC-N populations maintained on flat and nanotopographically-patterned MEAs (n = 48 wells per condition). (d) Measurement of the mean number of spikes recorded per electrode during individual network bursts from hPSC-N populations maintained on flat and nanotopographically-patterned MEAs (n = 48 wells per condition). (e) Measurement of network burst duration from hPSC-N populations maintained on flat and nanotopographically-patterned MEAs (n = 48 wells per condition). (f) Neuron migration speed on flat and nanotopographically-patterned substrates during the first 12 hours after seeding (n = 18 cells per condition). (g) Neurite outgrowth speed on flat and nanotopographically-patterned substrates during the first 12 hours after seeding (n = 11 cells per condition). (h) Human PSC-Ns on flat Nafion substrates immunostained using antibodies against dendrites (MAP-2) and axons (neurofilament) proteins. Inset shows detail of neurite orientation. (i) Human PSC-Ns on nanotopographically-patterned Nafion substrates immunostained using antibodies against dendrites (MAP-2) and axons (neurofilament) proteins. Inset shows detail of neurite orientation. The yellow arrow indicates the orientation of the underlying nanotopography. (j) Human PSC-Ns on flat Nafion substrates immunostained using antibodies against pre-(synaptotagmin) and post-synaptic (PSD-95) proteins. Inset shows detail of the staining associated with a single cell. (k) Human PSC-Ns on nanotopographically-patterned Nafion substrates immunostained using antibodies against pre-(synaptotagmin) and post-synaptic (PSD-95) proteins. Inset shows detail of the staining associated with a single cell. The yellow arrow indicates the orientation of the underlying nanotopography. (l) Pearson correlation coefficients calculated from analysis of scatter plots constructed from images of flat and nanotopographically-patterned hPSC-Ns as an indicator of the relative levels of pre- and post-synaptic marker co-localization (n = 10 images per condition). *p = 0.02, **p < 0.0005.