Biophysical models are increasingly used to gain mechanistic insights by fitting and reproducing experimental and clinical data. The inherent variability in the recorded datasets, however, presents a key challenge. In this study, we present a novel approach, which integrates mechanistic modeling and machine learning to analyze in vitro cardiac mechanics data and solve the inverse problem of model parameter inference. We designed a novel generative adversarial network (GAN) and employed it to construct virtual populations of cardiac ventricular myocyte models in order to study the action of Omecamtiv Mecarbil (OM), a positive cardiac inotrope. Populations of models were calibrated from mechanically unloaded myocyte shortening recordings obtained in experiments on rat myocytes in the presence and absence of OM. The GAN was able to infer model parameters while incorporating prior information about which model parameters OM targets. The generated populations of models reproduced variations in myocyte contraction recorded during in vitro experiments and provided improved understanding of OM’s mechanism of action. Inverse mapping of the experimental data using our approach suggests a novel action of OM, whereby it modifies interactions between myosin and tropomyosin proteins. To validate our approach, the inferred model parameters were used to replicate other in vitro experimental protocols, such as skinned preparations demonstrating an increase in calcium sensitivity and a decrease in the Hill coefficient of the force–calcium (F–Ca) curve under OM action. Our approach thereby facilitated the identification of the mechanistic underpinnings of experimental observations and the exploration of different hypotheses regarding variability in this complex biological system.

No abstract

Single cardiomyocytes are widely used for investigations of the cellular and molecular mechanisms of regulation and modulation of cardiac performance. Intact cardiomyocytes allow one to study in detail cell function avoiding the effects of extracellular matrix and neighboring cells. The most established protocols of cardiomyocyte isolation are based on the isolated heart perfusion using a Langendorff-apparatus or on intraventricular perfusion using a syringe. However, the yield of single cardiomyocytes obtained by these methods may be low due to the cell injury following non-uniform enzyme digestion of connective tissue in different heart chambers. Moreover, isolation of atrial cardiomyocytes is challenging because of their small size and complex geometric shape. Here we present a new protocol for simultaneous isolation of high quality cardiomyocytes from the atria, ventricular free walls and interventricular septum. The protocol is based on the combination of the Langendorff perfusion method with the intraventricular and intra-atrial injection technique taking into account the collagen content variation between the different heart chambers. Obtained cells demonstrate rod-shaped morphology, a clear and regular sarcomere striation pattern and rat-specific frequency-dependence of contraction and calcium transient parameters. Our protocol provides gentle cell isolation that increases the yield of single cardiomyocytes suitable for biophysical researches .

Cardiomyocytes contract keeping their sarcomere length (SL) close to optimal values for force generation. Transmural heterogeneity in SL across the ventricular wall coordinates the contractility of the whole-ventricle. SL heterogeneity (variability) exists not only at the tissue (macroscale) level, but also presents at the level of a single cardiomyocyte (microscale level). However, transmural differences in intracellular SL variability and its possible dependence on the state of contraction (e.g. end-diastole or end-systole) have not been previously reported. In the present study, we studied three aspects of sarcomere-to-sarcomere variability in intact cardiomyocytes isolated from the left ventricle of healthy guinea-pig: 1) transmural differences in SL distribution between subepi- (EPI) and subendocardial (ENDO) cardiomyocytes; 2) the dependence of intracellular variability in SL upon the state of contraction; 3) local differences in SL variability, comparing SL distributions between central and peripheral regions within the cardiomyocyte. To characterize the intracellular variability of SL, we used different normality tests for the assessment of SL distributions, as well as nonparametric coefficients to quantify the variability. We found that individual SL values in the end-systolic state of contraction followed a normal distribution to a lesser extent as compared to the end-diastolic state of contraction (∼1.3-fold and ∼1.6-fold in ENDO and EPI, respectively). The relative and absolute coefficients of sarcomere-to-sarcomere variability in end-systolic SL were significantly greater (∼1.3-fold) as compared to end-diastolic SL. This was independent of both the transmural region across the left ventricle and the intracellular region within the cardiomyocyte. We conclude that the intracellular variability in SL, which exists in normal intact guinea-pig cardiomyocytes, is affected by the contractile state of the myocyte. This phenomenon may play a role in inter-sarcomere communication in the beating heart.