Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia, with approximately 46 million patients worldwide. Yet to date, the exact molecular mechanisms underlying AF are still to be explored. The aim of this investigation was to use fast optogenetic ion channels and state-of-the-art atrial engineered heart tissue (EHT) to optimize an in vitro-model of AF.

We generated human atrial-like cardiomyocytes from induced pluripotent stem cells with an improved retinoic acid-based differentiation protocol, generated 24-well format EHTs and characterized their properties in terms of gene expression, function (force and action potentials) and drug responses. For optogenetic pacing, EHTs were transduced with one of three fast-gating light-sensitive cation channels (channelrhodopsins) PsCatCh, CheRiff, and Chronos via the adeno-associated virus vector AAV6. These three channelrhodopsins had faster (in)activation kinetics than those previously used in atrial EHTs. Expression level of the channelrhodopsins was evaluated via flow cytometry. The EHTs were paced with blue LED light as well as electrical stimulation over varying periods of time.

The EHTs showed an atrial gene expression pattern and canonical atrial-like drug responses to carbachol and 4-aminopyridine. Repolarization fraction (APD₉₀-APD₂₀)/APD₉₀ proved to be a good surrogate of “atrialness”. When paced electrically, EHTs could follow the impulses up to a frequency of 7 Hz. Depending on the expression level of the respective channelrhodopsin, EHTs could follow optical pacing to similar frequencies in short- and long-term experiments. They also exhibited structural and electrophysiological remodeling over time.

Our optogenetic pacing approach was successfully implemented in atrial EHTs and important technical questions could be successfully addressed. The model will now be used to answer the question whether electrophysiological remodeling resulting from different pacing-based pre-conditioning patterns could evoke a propensity to arrhythmia.