Auriane Ernault

Atrial fibrillation (AF) is the most common cardiac arrhythmia and has a large medical and societal impact. It is associated with decreased quality of life, increased risk of stroke and mortality. Risk factors of atrial fibrillation include obesity, which is associated with an increased volume of epicardial adipose tissue. Although observational studies document the association between epicardial adipose tissue volume and arrhythmias, the underlying electrophysiological mechanisms are unknown. In this thesis, we identified mechanisms by which epicardial adipose tissue modulates cardiac arrhythmogenesis.

In Chapter 2, we reviewed literature and summarized evidence showing that epicardial adipose tissue accumulation is associated with cardiac conduction slowing. We discussed various structural and functional pathways by which epicardial adipose tissue can facilitate arrhythmias. Adipose tissue infiltration between myocytes creates an anatomical obstacle to cardiac excitation, which delays activation and increases heterogeneity in the myocardium, thus facilitating the genesis of life-threatening reentrant arrhythmias. Furthermore, epicardial adipose tissue cells can couple to cardiomyocytes and induce conduction slowing. Finally, epicardial adipose tissue secretes adipokines that can modulate cardiac arrhythmogenesis by modifying ion channels function and expression, by altering electrical coupling between cardiomyocytes, and by stimulating fibrosis. All of those mechanisms facilitate reentrant arrhythmias and triggered activity.

Because there is no fascia separating epicardial adipose tissue from the myocardium, we hypothesized that epicardial adipose tissue secretes adipokines which induce electrical remodeling of the myocardium, facilitating arrhythmias. In Chapter 3, we showed results of the experimental test of this hypothesis. We showed that epicardial adipose tissue secretome from atrial fibrillation patients induces electrical remodeling of adjacent myocardium by reducing Kcnj2 expression, depolarizing the resting membrane of cardiomyocytes and decreasing electrical coupling. This resulted in conduction slowing and increased conduction heterogeneity in a neonatal rat ventricular myocytes model. On the other hand, secretome from subcutaneous adipose tissue did not lead to such changes. Using an in silico model of human left atrium we demonstrated that the electrophysiological changes of the myocardium induced by epicardial adipose tissue secretome facilitate sustained reentrant arrhythmias.

Next, we aimed to identify the components in the epicardial adipose tissue secretome responsible for this electrical remodeling. We hypothesized that epicardial adipose tissue induces arrhythmogenic conduction slowing by secreting extracellular vesicles and their microRNA cargo. In Chapter 4, we showed that epicardial adipose tissue secretome contains higher extracellular vesicle concentration than secretome of subcutaneous adipose tissue. We further demonstrated that epicardial adipose tissue secretome-derived extracellular vesicles contain hundreds of microRNAs, some of them predicted to be related to cardiac electrophysiology through regulation of resting membrane potential and potassium channels activity. We reported that miR-1-3p and miR-133a-3p are more expressed in epicardial adipose tissue than in subcutaneous adipose tissue. Finally, we showed that overexpression of miR-1-3p or miR-133a-3p in neonatal rat ventricular myocytes recapitulates the same heterogenous conduction slowing effect as epicardial adipose tissue secretome in association with decreased expression of Kcnj2, as demonstrated in Chapter 3. Thus, the microRNA cargo of extracellular vesicles from epicardial adipose tissue secretome can be important players of epicardial adipose tissue secretome-induced arrhythmogenicity, and may become targets for therapy.

Atrial fibrosis plays an important role in the development and persistence of atrial fibrillation by promoting reentry. Primary cilia have been identified as a regulator of fibroblasts activation and extracellular matrix deposition. In Chapter 5, we therefore tested the hypothesis that selective reduction of primary cilia in fibroblasts causes increased fibrosis and facilitates reentry. We genetically disrupted the formation of primary cilia in neonatal rat ventricular fibroblasts and examined the consequences of reduction of ciliated fibroblasts on conduction and tissue remodeling by performing electrical mapping, microelectrode recording, and gene expression measurements. We show that the disruption of cilia formation in fibroblasts is associated with enhanced extracellular matrix gene expression, conduction delay and increased risk of spontaneous reentry in cardiomyocyte-fibroblast co-cultures. Therefore, prevention of cilia loss can be a novel target for prevention of arrhythmias.

In the previous chapters we have cultured neonatal rat ventricular myocytes on multielectrode arrays as a method to assess the electrophysiological characteristics and their long term changes after exposure to epicardial adipose tissue secretome or increased extracellular matrix. However, the interpretation of the local electrograms is not unequivocal. In Chapter 6, we used an in silico model of a multielectrode array to validate the interpretation of extracellular field potentials recorded from neonatal rat ventricular myocytes. We show that field potential duration can be used as a measure for action potential duration. We also show that local extracellular action potential (LEAP) measurements are not resulting from high coupling of cardiomyocytes on the array, but from local depolarization of cardiomyocytes which leads to monophasic action potential measurement on the site of LEAP. We also show that the LEAP is not a reliable method to capture action potential morphology.

In Chapter 7, we interpret and discuss the relation between the key findings of this thesis. We have identified several extra-myocardial factors that alter conduction through paracrine or juxtracrine remodeling. These factors provide new inroads for alternative antiarrhythmic therapy for example by prevention of cilia loss of cardiac fibroblasts, by reduction of extracellular vesicles release by epicardial adipose tissue, by targeted anti-miR strategies, or by reducing epicardial adipose tissue volume by body weight reduction.