Investigation of the structural and molecular substrate of atrial fibrillation

Atrial fibrillation (AF) is the most common sustained arrhythmia worldwide and a frequent cause of hospitalization. Moreover, it represents one of the most frequent complication following cardiac surgery with an incidence of around 30% and an important predictor of patient morbidity. The exact pathophysiological mechanisms responsible for the onset and perpetuation of AF are not completely understood. However, clinical and experimental insights on the factors causing AF have suggested that atrial fibrillation is a multi-factorial phenomenon. Atrial fibrillation is characterized by a highly complex and irregular electrical activation of the atrial tissue, which is the manifestation of diverse abnormalities (electrical, structural, metabolic, neurohormonal, and/or molecular alterations) in diverse pathological conditions. In particular, it has been shown that fibrosis, a phenomenon in which extracellular matrix (ECM) components, mainly fibrillar collagen, accumulate between cardiomyocytes, leads to the inhomogeneous atrial electrical conduction typical of fibrillation. Recent studies have suggested that the deregulation of gene expression may act as a molecular mechanism of arrhythmogenesis. In particular, miRNAs, a new class of non-coding RNAs have rapidly emerged as one of the key players in the gene expression regulatory network, so variations in their expression levels may constitute a pathway for the arrhythmia-induced atrial remodeling. The present study aims to investigate the structural and molecular features of atrial tissue, with particular attention to fibrosis, which may be involved in the formation of a pro-arrhythmic substrate. By using both histological and advanced microscopy techniques, intramural fibrotic content and 3D collagen network properties were determined in atrial samples, collected during cardiac surgery in patients who developed or not AF. The quantitative analysis indicated a general decrease of collagen content from the outer (the epicardium) to the inner (the endocardium) myocardial wall, in the overall patient population. However, AF patients presented higher fibrotic values compared to sinus rhythm (SR) patients in the deeper myocardial layers, thus supporting the hypothesis that an accumulation of fibrotic tissue within the myocardial wall may represent an important structural contributor in the pathophysiology of AF. In addition to a quantitative assessment, collagen properties such as fibers orientation (degree and anisotropy) and scale dimension, were determined by non-linear optical microscopy techniques. The analysis revealed that in SR patients collagen network showed a fine architecture characterized by thin fibrils with changing angles and directions compared to AF, where fibers tended to pack-up in larger bundles of defined directions. A quantitative analysis of the 3D collagen network features, throughout the atrial wall, revealed that fibers orientation and scale dimension changed along tissue depth in both SR and AF patients, with larger values of orientation and fiber changes in AF tissues. These results highlight the spatial rearrangement and thickening of the 3D collagen network in AF patients, suggesting its possible role in the maintenance of the arrhythmia. Numerous evidence indicated that also an altered regulation of gene expression may play an important role in the mechanisms of atrial remodeling which underlie AF. In this perspective, the expression pattern of some miRNAs known to target different genes involved in diverse mechanisms that underlie AF was evaluated. A panel of miRNAs (miR-1, miR-133a/b, miR-30c, miR-29a/b, miR-208a/b, miR-328, miR-499, miR-590 and miR-21), principally involved in the formation of a pro-arrhythmic substrate, was selected after an accurate review of the literature and analyzed by RT-qPCR, in AF patients versus SR individuals. To accurately determine the levels of analyzed miRNAs, their expression data are usually normalized relatively to endogenous and/or exogenous reference genes. To date, no general agreement between different normalization strategies has been found, in particular in cardiac tissue, for the study of AF. For these reasons, a preliminary study aiming to establish the best endogenous reference genes for miRNAs data normalization was performed. Specifically, different well-established analysis tools such as NormFinder, GeNorm, BestKeeper and ∆Ct method, were applied on five commonly used endogenous reference transcripts such as 5S, U6, SNORD48, SNORD44 and miR-16. The suitable reference gene obtained, SNORD48, was applied for miRNAs data normalization. Our findings revealed that miRNAs expression levels were different in AF compared to SR patients. MiR-208a and miR-208b displayed statistically significant differences between the two populations. To investigate possible relationships between miRNAs expression levels and the fibrotic content a correlation measurement was also performed. Our analysis revealed that miR-21 and miR-208b were close to a significant correlation with fibrosis. In conclusion, this work introduced new techniques and implemented new methods of analysis for the study of the substrate of AF. In particular, the results obtained with this multiscale approach, from structural to molecular level, exacerbated the role of fibrosis as a critical contributor in the formation of a pro-arrhythmic substrate. Nonetheless, further studies are needed for a better understanding of the ways in which structural, molecular and also cellular remodeling may alter the impulse propagation in the myocardium.