Progressive stretch enhances growth and maturation of 3D stem-cell-derived myocardium

Bio-engineered myocardium has great potential to substitute damaged myocardium and for studies of myocardial physiology and disease, but structural and functional immaturity still implies limitations. Current protocols of engineered heart tissue (EHT) generation fall short of simulating the conditions of postnatal myocardial growth, which are characterized by tissue expansion and increased mechanical load. To investigate whether these two parameters can improve EHT maturation, we developed a new approach for the generation of cardiac tissues based on biomimetic stimulation under application of continuously increasing stretch. Methods: EHTs were generated by assembling cardiomyocytes derived from human induced pluripotent stem cells (hiPSC-CM) at high cell density in a low collagen hydrogel. Maturation and growth of the EHTs were induced in a custom-made biomimetic tissue culture system that provided continuous electrical stimulation and medium agitation along with progressive stretch at four different increments. Tissues were characterized after a three week conditioning period. Results: The highest rate of stretch (S3 = 0.32 mm/day) increased force development by 5.1-fold compared to tissue with a fixed length, reaching contractility of 11.28 mN/mm². Importantly, intensely stretched EHTs developed physiological length-dependencies of active and passive forces (systolic/diastolic ratio = 9.47 ± 0.84), and a positive force-frequency relationship (1.25-fold contractility at 180 min-1). Functional markers of stretch-dependent maturation included enhanced and more rapid Ca2+ transients, higher amplitude and upstroke velocity of action potentials, and pronounced adrenergic responses. Stretch conditioned hiPSC-CMs displayed structural improvements in cellular volume, linear alignment, and sarcomere length (2.19 ± 0.1 µm), and an overall upregulation of genes that are specifically expressed in adult cardiomyocytes. Conclusions: With the intention to simulate postnatal heart development, we have established techniques of tissue assembly and biomimetic culture that avoid tissue shrinkage and yield muscle fibers with contractility and compliance approaching the properties of adult myocardium. This study demonstrates that cultivation under progressive stretch is a feasible way to induce growth and maturation of stem cell-derived myocardium. The novel tissue-engineering approach fulfills important requirements of disease modelling and therapeutic tissue replacement.


Flow cytometry
The cell suspension used for EHT generation was tested for purity of hiPSC-derived cardiomyocytes by FACS analysis. For intracellular staining of cardiac troponin T (cTnT) and alpha-actinin, 0.5×10 6 dissociated cells were suspended in 15 mL 4% paraformaldehyde and fixed for 15 min. After centrifugation (200 xg, 5 min, 4 °C) cells were permeabilized in PBS with 1% Triton X-100 and washed twice again with PBS containing 1% BSA (Sigma). For determination of cardiomyocyte volumes, the cell suspension of dissociated EHTs was subjected to FACS analysis. EHTs were mechanically reduced to small pieces which were incubated on a thermostated shaker (20 min, 37 °C) with dissociation solution (collagenase II 2,2 mg/mL, Sigma Aldrich and TrypLE Select Enzyme, Gibco). After further mechanical disruption by pipetting, undissolved tissue strips were removed, and the cell suspension was stained for cardiac troponin T (cTNT) as described. Data shown are representative of five independent experiments.
To assess calcium handling, EHTs within BMCCs were loaded with 5 µmol/L Fluo-4, AM (Invitrogen) under electromechanical stimulation at 1 Hz for 30 minutes at 37 °C, followed by an additional 15 min washout period after medium exchange. Videos were acquired on an inverted microscope during continuous stimulation at a rate of 50 frames/s, using a Rolera EM-C2 camera (Teledyne QImaging, Canada), and data were acquired with Micro-Manager Studio Version 1.4 (University of California San Francisco, USA). Quantitative data were extracted with ImageJ (National Institute of Health, USA), and calcium dynamics were analyzed with WinEDR v3.8.6 (University of Strathclyde, UK). For each experiment, several regions of interest were selected for calculation of the mean fluorescence intensity. The signal was normalized to the diastolic background fluorescence. Calcium transient duration, amplitude, and calcium decay rate were measured.

Simulated hypoxia
For induction of tissue hypoxia, the standard condition of medium agitation (tilting of BMCCs by a 12° angle at 1 Hz) was stopped for a 2 min interval. Electrical stimulation at 1 Hz and contraction force recording was continued throughout.

Isometric contraction performance
Contractility of the EHTs were determined under isometric conditions in horizontal organ baths (Mayflower, Hugo Sachs Elektronik, Germany) as previously described [1]. In brief, EHTs were attached to isometric force transducers (F30, HSE) with preload adjusted to 1 mN, and continuously perfused with oxygenated Tyrode solution (1.8 mmol/L CaCl2 and 23 mmol/L bicarbonate, pH 7.4) for at least 10 min prior to any measurement. Electrical stimulation (1 Hz, 3 ms pulse width, 2-fold stimulation threshold) was initiated immediately.
The maximum twitch force was determined under optimum preload. For determination of length-tension relationship, tissues were distended in steps of 0.25 mm. Baseline (slack length) was defined as the longest extension that produced no diastolic force. Passive and active forces were recorded continuously and analyzed 15 seconds after each distension step.
The force-frequency relationship was determined by the stepwise increase of stimulation frequency from 1 Hz to 6 Hz. The elastic modulus was calculated from the slope of diastolic force at the distension preceding the condition of the maximum contractility. The slope of diastolic force was related to the length and actinin + cross-section area of the tissue.

Isoproterenol response
Isoproterenol was dissolved in Tyrode solution and added to the organ bath at a 1:1000 v/v dilution. The contraction force was continuously recorded. Drug-induced changes in the generated force, relaxation rate, and contraction duration were determined.

Action potential recording
Action potentials (APs) were recorded with standard sharp microelectrodes in intact EHTs after 21 days of culture, as previously described [2]. Tissues were mounted at relaxed length  Table S2.

Immunostaining
After washing three times with PBS, whole tissues were prepared for cryosectioning. First, EHTs were placed at 4 °C overnight in 1 mL of 30% sucrose solution in PBS. The next day, we embedded the EHTs in a 7.5% gelatin, 30% sucrose solution. A freezing bath was prepared by dropping several small pieces of dry ice into a bath of isopentane to reach a temperature between -50 ° and -30 °C. The block of gelatin was immersed into the cold bath and allowed to freeze for 2 min. Sections were cut for immunostaining in a standard cryostat.
The sectioned tissues were washed with PBS for 30 minutes, permeabilized with 1% Triton X-100 in PBS for 60 min, and then incubated in blocking solution (3% bovine serum albumin and 1% fetal calf serum in PBS) for 1 hour. The primary monoclonal anti-α-actinin antibody (sarcomeric, 1:100, Sigma-Aldrich) in antibody dilution buffer (Supplement Table S1) was incubated overnight 4 °C. The next day, samples were incubated with goat anti-mouse IgG (H+L) highly cross-adsorbed secondary antibody, Alexa Fluor 546 (Invitrogen), at 1:100 in antibody dilution buffer for 2 hours. Tissues were washed three times for 10 min in PBS and subsequently incubated with DAPI (2 µmol/L, Invitrogen), and wheat germ agglutinin (WGA) conjugated to CF633 (BioTium CF) at 40 µg/mL overnight at 4 °C. Sections were mounted with Fluoromount-G (Invitrogen).  B Figure S4. The threshold for EHT pacing decreases during culture. The stimulation current required for electrical excitation of EHTs progressively declined during the first 3 weeks of culture. No difference was detected between the various stretch conditions. The graph displays exemplary data from EHTs exposed to different stretch conditions (S0 -S3, n = 6).