Atrial fibrillation in a dish: insights into atrial arrhythmogenesis from induced pluripotent stem cell-derived cardiomyocytes

This editorial refers to ‘Human iPSC modeling of a familial form of atrial fibrillation reveals a gain of function of I_f and I_CaL in patient-derived cardiomyocytes’, by P. Benzoni et al., pp. 1147–1160.

Atrial fibrillation (AF) is a common arrhythmia with increasing prevalence. The underlying causes of AF maintenance and progression are poorly understood.^1,2 This might partly explain the low efficacy of current antiarrhythmic drugs (AADs) and ablation therapy, which are also associated with side-effects and proarrhythmia (particularly with AADs). Thus, there is an urgent unmet need for novel therapeutics, with the hope that an improved understanding of AF pathophysiology will help to develop drugs with improved efficacy and safety.^3,4

Atrial remodelling is a key element of AF pathophysiology and a major contributor to the progression of AF to more persistent forms.² While there is ample evidence that the fast atrial rates during AF contribute to the creation of an AF-maintaining atrial substrate, the mechanisms of AF initiation in the context of atrial cardiomyopathy are less clear.^2,5 In this context, genetic causes of AF are instrumental in the dissection of mechanisms predisposing to AF induction and re-initiation. Most importantly, the advent of the human induced pluripotent stem cell (iPSC) technology and the ability to differentiate iPSCs from patients with a genetic defect into cardiomyocytes (iPSC-CMs)⁶ permit to model and study the pathophysiology of human arrhythmias (Figure 1). Coupled with advances in genome editing and next-generation sequencing technologies, iPSC-CMs are a powerful technology to accelerate the investigation of molecular mechanisms for cardiac diseases.⁷ Patient-specific iPSC-CMs also provide a model for personalized medicine, by enabling generation of ‘patient in a dish’ models of disease.⁷

Figure 1.

The ability to generate induced pluripotent stem cells from somatic cells of patients and to differentiate them into cardiomyocytes (iPSC-CMs) allows to create ‘disease-in-a-dish’ models and generate new insights into the molecular mechanisms of inherited cardiac diseases and arrhythmia syndromes, including atrial fibrillation. Human iPSC-CMs hold the promise to become an enhancing technology to accurately predict disease mechanisms, facilitate drug discovery, and personalize treatment. PPI, protein–protein interaction.

Benzoni et al.⁸ used patient-derived iPSC-CMs to investigate the triggering cellular mechanisms of a familial form of AF. This is the first report of a human cellular model of AF from siblings with untreatable persistent AF. It identified peculiar functional alterations in CM electrophysiology in the common genetic background of the patients leading to a cardiac substrate more prone to develop and accommodate arrhythmias under stressful conditions. The authors found that iPSC-CMs from AF patients have longer action potential (AP) duration, larger L-type Ca²⁺-current (I_Ca,L) amplitude, along with a greater Ca²⁺-transient amplitude and an increased sarcoplasmic reticulum Ca²⁺ load. The pacemaker current I_f was enhanced, with no change in K⁺-currents. These data suggest that enhanced I_f-mediated automaticity and prolonged AP with subsequent increases in the dispersion of repolarization, potentially causing reentry-promoting conduction block, might provide the arrhythmogenic trigger and substrate, respectively, for AF in these individuals.

The results of the present study confirm some key previous findings in adult cardiomyocytes from AF patients. Indeed, Ca²⁺-transient amplitude is higher in paroxysmal AF vs. sinus rhythm patients.⁹ Cells from patients with paroxysmal AF and with systolic heart failure, who are prone to AF, show increased Ca²⁺ loading and a higher susceptibility to proarrhythmic diastolic Ca²⁺ releases (Ca²⁺ leak) and delayed afterdepolarizations (DADs).^9,10 Despite abnormal Ca²⁺ handling, iPSC-CMs from AF patients in the present study did not exhibit more frequent DADs. This might have been precluded by the enhanced automaticity in both cell lines. I_K1, which is lacking in these cells, might have differently affected membrane potential stability in the atrial cardiomyocytes from patients with paroxysmal AF and systolic heart failure. Although an enhanced I_f current has not previously been reported as a cause of AF, an increase in the mRNA and protein levels of the I_f-channel subunits HCN2 and HCN4 was reported in atria of heart failure patients.¹¹ Thus, it is conceivable that enhanced I_f might provide a trigger for AF induction in some patients.

Despite the lack of a state-of-the-art differentiation to atrial-like iPSC-CMs¹² in the present study, results are encouraging that iPSC-CMs can help to identify potential AF mechanisms. The technology could be exploited to reveal viable therapeutic targets, improve patient stratification, and predict the therapeutic response of AF patients to a particular AAD (Figure 1). Nevertheless, iPSC-CMs have several important limitations. First, the cells have a rather immature phenotype, characterized by spontaneous beating, likely due to the presence of early-developmental currents and overall ionic fingerprint that differs significantly from that in native CMs and more closely resembles foetal CMs. iPSC-CMs have immature Ca²⁺-handling machinery and a rudimentary ultrastructure, e.g., lack of transverse tubules. Significant efforts are underway to enhance the structural and functional maturation of iPSC-CMs, including manipulation of growth conditions, electrical and/or mechanical stimulation, and three-dimensional systems that provide an environment closer to in vivo heart development.¹³ Another challenge lies in the iPSC-CM phenotypic variability, whereby AP shapes and Ca²⁺ signals have been shown to vary widely even within a single cell line. IPSC-CM variability obviously also stems from differences in genetic information of donors. This is an important feature of iPSC-CMs, as it allows accounting for patient phenotype variation and observing a broad spectrum of responses to perturbations (including drugs). However, linking the varied phenotypes and responses to varied genomic, proteomic, or ionic mechanisms is challenging. Machine learning approaches are being utilized to find correlations between phenotype and genetic, electrophysiological, and contractile features, but mechanistic conclusions regarding the molecular underpinnings are rather limited. Mechanistic models can provide a complementary quantitative framework to address this. Recently, a population of human iPSC-CM models incorporating experimentally observed variations in ionic current recapitulated physiological ranges of variability.¹⁴ This high-throughput modelling and simulation approach, which is amenable to both statistical analysis and mechanistic investigations, could be a critical means to assess how varied genetic background confers resilience or increased arrhythmia susceptibility to any individual model variant.

In conclusion, the use of human iPSC-CMs for modelling AF in a dish, as illustrated in this study, holds promise to accelerate and complement approaches aimed at elucidating the underlying cellular mechanisms of AF, and identifying novel mechanism-based therapies that could be individualized to each patient. The study by Benzoni et al.⁸ validates the potential causal contribution of increased I_f, prolonged AP (via increased dispersion of refractoriness and conduction block), and enhanced sarcoplasmic reticulum Ca²⁺ load (and Ca²⁺ leak) to AF promotion. Further studies employing iPSC technology in affected individuals will help to expand our understanding of AF pathophysiology and could ultimately lead to the discovery of novel anti-AF targets and effective therapeutic approaches.

Conflict of interest: Dr E.G. has no conflicts of interest to disclose. Dr D.D. is member of Scientific Advisory Boards of OMEICOS Therapeutics, Acesion Pharma, and Sonofi and received speaker’s fees for educational lectures from Boston Scientific, Novartis, and Bristol-Myers Squibb. His laboratory executed research contracts for OMEICOS.

Funding

The author’s research work is funded by the National Institutes of Health (NIH) Stimulating Peripheral Activity to Relieve Conditions grant (1OT2OD026580-01 to E.G.); and NIH grants (R01HL131517 to E.G. and D.D.; R01HL136389 to D.D.; R01HL089598 to D.D.); the German Research Foundation (Do 769/4-1 to D.D.); American Heart Association (15SDG24910015 to E.G.); and UC Davis School of Medicine Dean’s Fellow award to E.G.

The opinions expressed in this article are not necessarily those of the Editors of Cardiovascular Research or of the European Society of Cardiology.