Human Engineered Heart Tissues Reproduce Trigger‐ and Substrate‐Driven Mechanisms of Ventricular Tachycardia

Figures

• 

  FIGURE 1. Reproducible fabrication of EHTs using a modified milliPillar platform. (A) Modified milliPillar molds were 3D‐printed in temperature‐resistant resin and used to cast PDMS. (B) Individual modules were attached to multi‐well dishes, increasing flexibility and compatibility with standard electrophysiology and cell culture workflows. (C) Representative brightfield image of an EHT 1 month after plating. (D) Confocal section stained for DAPI, α‐actinin, and vimentin; magnified insert shows longitudinal cardiomyocyte alignment with interspersed fibroblasts (vimentin‐positive). (E) Confocal section stained for DAPI, α‐actinin, and troponin T (TnT). Magnified insert reveals aligned sarcomeric banding. (F) An IonOptix C‐Pace insert was used to deliver field electrical stimulation. (G) Contractility was assessed from 200‐fps brightfield videos under stimulation at 37°C in maturation media; the final 15 s of each pacing frequency plus a 5 s padding were analyzed with BeatProfiler. A positive force‐frequency relationship was observed in all EHTs (H). (I–N) Quantifications are shown as box plots (median, interquartile range (IQR), and 1.5X IQR whiskers) overlaid with scatter points for N = 12…

• 

  FIGURE 2. Optical mapping setup and EHT electrophysiology. (A) Schematic representation. “Created in BioRender. Altamirano, F. (2026) https://BioRender.com/harfohp”. (B) Photograph showing LED, mesoscope, camera, and optics. (C) 3D rendering of STL files for the imaging stage: a perfusion heater holder and epoxy‐encapsulated heating elements maintain temperature during recording, while EHT holders and pocket minimize motion and flotation. An adjustable slider with two platinum wires provides point‐ or field‐electrical stimulation. (D) Top‐down view showing milliPillar module positioning for a point‐electrical stimulation experiment. (E) Ensemble voltage (top) and Ca2+ (bottom) activation maps from a representative EHT paced at 2 Hz. Purple circles denote ROIs shown in panel (F) showing normalized voltage (blue) and Ca2+ (orange) signals illustrating consistent propagation, with voltage preceding Ca2+. (G–I) Dominant frequency (DF) and ensemble APD80, CaD80, and activation maps at 1 and 3 Hz. (J–M) Quantifications of APD80/CV for voltage and CaD80/tau for Ca2+ signals reveal rate‐dependent restitution curves. Data are expressed as cycle length (CL = (1/frequency) × 1000 ms). N = 11–18 EHTs…

• 

  FIGURE 3. Generation of VT‐like condition in EHTs. High‐dose E‑4031 (500 nM) combined with low extracellular K+ and Mg2+ (aLQTS) created a proarrhythmic substrate that emerged within approximately 1 min and persisted for the duration of the observation period (<5 min). (A) Diagram of contractility and optical mapping assessments under baseline and aLQTS conditions. “Created in BioRender. Altamirano, F. (2026) https://BioRender.com/btrdquu”. (B) Representative active force trace and Poincaré plot from a control (CTL) EHT at 1 Hz pacing demonstrates 1:1 capture and consistent beat‐to‐beat synchrony. (C) Representative active force trace and Poincaré plot from an EHT following acute exposure to low K+/Mg2+ modified Tyrode with 500 nM E‐4031 (aLQTS condition) shows transient increases in beating rate and beat‐to‐beat variability. (D–K) Summary quantifications of N = 19 EHTs (4 batches) recorded under CTL and then aLQTS conditions. All data are shown as box plots (median, IQR, and 1.5X IQR whiskers) with overlaid scatter points. Paired data were compared using a two‐sided paired Wilcoxon signed‐rank test because several metrics showed non‐normal paired differences.

• 

  FIGURE 4. aLQTS condition prolongs APD80, increases dispersion, and promotes wavebreak, reentry, and a VT‐like phenotype in EHTs. Multiple dual voltage and Ca2+ optical mapping recordings were obtained per condition. Fraction of recording time spent above 120 (A) and 140 (B) bpm for EHTs paced at 1.0 Hz. Five EHTs from iPSC line 1 were paced at 1.5 Hz; these recordings were pooled with the 1.0 Hz data because the results were comparable. Shapiro‐Wilk indicated non‐normality for several distributions; unpaired two‐sided Mann‐Whitney U tests were applied (CTL N = 27, aLQTS = 28, from 4 batches for iPSC 1 and 2 batches for iPSC2s). Box‐and‐whisker plots show median and interquartile range; dots are individual EHTs. C. Voltage (left) and Ca2+ (right) kymographs from a CTL (top) and aLQTS (bottom) recording showing simultaneous activation under pacing in the CTL and progressive loss of capture, beat acceleration, and slowed CV under aLQTS (Video S5 and S6). D‐F. Heatmaps showing pixel‐wise changes across beats for APD80, DI, and dV/dt from a representative EHT under CTL (left) and aLQTS (right) conditions. APD80 becomes prolonged and spatiotemporal localized heterogeneity increases under aLQTS…

• 

  FIGURE 5. Heterogeneous EAD behavior and propagation in an aLQTS. Top panel, Simultaneous dual optical mapping of transmembrane voltage (black) and intracellular calcium (Ca, green) recorded from four regions of interest (ROI 1–4) of aLQTS EHT paced at 1 Hz (field stimuli timestamps indicated by orange dashed lines and arrowheads). Traces are min‐max normalized to the first‐99th percentile range (0‐1). EADs were absent in control and emerged after aLQTS, displaying a spectrum of behaviors across ROIs and beats: subthreshold EADs that fail to depolarize the tissue (open circles), attenuated EADs that propagate locally but decay before reaching neighboring regions (red‐purple stars), and EADs that give rise to triggered activity propagating across the tissue (red star into open stars). Black arrows between ROI rows denote the direction of propagation. Blue shaded windows (1–3) mark the segments expanded on the bottom. Bottom panel, Zoom 1–3: magnified Voltage (black) and Ca (green) traces from the four ROIs for each highlighted event, together with a schematic of the EHT indicating the four ROI positions (o1–o4), the propagation direction (black arrow), and, when applicable, the site of…

• 

  FIGURE 6. Phase singularities and rotor dynamics in aLQTS‐condition EHTs. (A) Voltage (left) and Ca2+ (right) kymographs with PS tracks overlay from a second representative aLQTS recording with VT‐like phenotype during field pacing at 1 Hz. Continuous propagation was analyzed using kymographs, and areas with signal fragmentation were masked (green) and resolved with ROI‐based trace plots (see below). (B) Pixel‐wise longitudinal APD80 heatmap demonstrates APD gradients between EHT heads and shaft. Prior to reentry, APD dispersion within the heads progressively increased as DI progressively shortened during spontaneous beating acceleration (C). (D) Areas with reduced excitability (dV/dt minima) increased over time near the EHT head‐neck joints and preceded the formation of PS/rotors. (E) Activation time, APD80, and dV/dt maps for beats 12–14 show a progressive increase in activation time and circular propagation around the right EHT head. APD80 increased over time and progressively generated high‐low APD areas around the right EHT head, linking the initial slow and circular propagation to a reduction in dV/dt that preceded conduction block, wavebreak, and reentry. (F) Dominant frequency maps…

• 

  FIGURE 7. Highly complex activation patterns in a subset of aLQTS‐condition EHTs exhibit VF‐like fragmentation. (A) Voltage (left) and Ca2+ (right) kymograph with overlaid PS tracks from an aLQTS EHT. (B) Magnified view highlights the spatiotemporal heterogeneity of voltage and Ca2+ signals (Video S9a). (C) Voltage signal and phase map with highlighted voltage wavelets reveal abnormal activation coinciding with a PS and two simultaneous wavelets and signal heterogeneity. (D) Quantification of simultaneous wavelet and PS counts over time in the representative aLQTS EHT demonstrates VF‐like presence of multiple wavelets.

• 

  FIGURE 8. Cryoinjury (Cryo) in EHT heads generates a spatial substrate for reentry. (A) Dual voltage and calcium optical mapping was performed before and after Cryo. Cartoon showing experimental design. Cryo was achieved using a liquid nitrogen‐chilled 1 mm copper pipe. The tip was flattened to cover the surface between the silicone edge and the center of the left EHT head. One edge of the frozen pipe was positioned on the silicone and then gently tilted toward the EHT head until a freezing halo covering around 50% of the head was visible. Optical mapping was performed thereafter within a 5 min period. (B) Voltage and Calcium signals were denoised and normalized from 0 to the 70th percentile 70% to identify damaged areas. The left panel shows the tissue before and after (CTL and Cryo, respectively). (C) A combination of signal‐to‐noise ratio and peak amplitude was used to evaluate injured areas. 10/10 cryoinjured EHTs showed a damaged area within the left head. (D) Cryo promoted tachyarrhythmias characterized by events with HR >120 bpm during a 1 Hz (60 bpm) point stimulation pacing delivered on the right head. (E) Cryo increased (phase singularities) PS compared to CTL (Total 62 events; 1…