Mechanical characterisation of live atrial tissue slices from patients with atrial fibrillation or sinus rhythm

Fibrosis represents a primary structural remodelling process occurring in atrial fibrillation (AF); however, its impact on atrial mechanics remains poorly understood. To address this knowledge gap, we developed an integrative method to simultaneously study passive and active mechanics, as well as the structure of atrial tissue slices. We hypothesized that there is a correlation between slice stiffness, active force generation, and tissue composition, and that stiffness and fibrosis are higher in tissue obtained from patients in AF, compared to sinus rhythm (SR).
Atrial slices (400 μm thick, 3-5 mm long and wide) from patients in SR or AF were kept contracting auxotonically in biomimetic chambers, while monitoring systolic and diastolic forces. A tensile test including low (2.4±0.25% elongation, mean±SEM) and high strain (additional elongation up to 15±1.7%) was performed in the presence of 30 mM 2,3-butanedione monoxime (BDM), as well as 2h after its removal. Arrhythmia susceptibility was examined by perfusing the tissue with isoproterenol (0.3 µM) and by additionally applying acute stretch up to 30%. Samples were subsequently fixed at their control length with 4% paraformaldehyde, stained with Picrosirius red, imaged using an automated slide scanner in brightfield mode (Zeiss Axioscan), and segmented to calculate the collagen fractional volume.
Our model showed sustained steady state contraction amplitudes, both for SR (N=7 patients, n=42 slices) and AF slices (N=3, n=11); active force generation showed no systematic differences between groups. No significant difference in diastolic stiffness were observed in the presence or absence of BDM. Passive stiffness during low strain was higher in SR slices (41.6±5.1 kPa) than AF (19.5±2.7 kPa). During high strain, this difference disappeared. Stiffness during low strain correlated with fractional volume and median thickness of collagen in individual samples. Comparing SR and AF group data, collagen thickness was higher in SR than in AF, while overall fractional volume was not significantly different. SR samples showed a stronger isoproterenol response in active force production (fold-change 9.9±2.5 for SR and 1.6±0.3 for AF slices), however no significant differences in arrhythmogenesis were observed between SR and AF slices upon application of isoproterenol, or upon 30% stretch (28.3% arrhythmias in SR; 25% in AF slices).  
At low strain levels, SR slices exhibited greater stiffness compared to AF slices, and this was correlated to collagen thickness. Upon increasing strain levels, AF slices showed a comparatively higher increase in stiffness than SR. This is counterintuitive in several ways: one would have expected higher fibrosis and stiffness in AF compared to SR, and a correlation between collagen levels and stiffness at high rather than low strain. That said, SR tissue samples were from patients who underwent cardiac bypass surgery, so the assumption that this is ‘healthier’ tissue than that from AF patients may be unfounded. Our data further suggest that tissue mechanical properties of atrial myocardium are not solely determined by collagen. To further elucidate these aspects, future research will focus on identifying the specific tissue components that contribute to passive mechanical properties of atrial tissue, ideally including atrial tissue from healthy donor tissue.