Cardiac mechano‐electric crosstalk: multi‐scale observations, computational integration, and clinical implications

Almost all aspects of cardiac electrical function are mechano-sensitive. From excitation of the sinoatrial node (MacDonald & Quinn, 2021), to conduction through the Purkinje network (Dominguez & Fozzard, 1979), atria (Solti et al., 1989) and ventricles (Zabel et al., 1996), and recovery during repolarisation (Quinn, 2015), mechano-electric interactions are critical for autoregulatory maintenance of the heart's physiological activity, and when disturbed can have deadly consequences (Quinn & Kohl, 2021). This Special Issue of The Journal of Physiology is focused on emerging aspects of Cardiac Mechano-Electric Crosstalk inspired by topics highlighted at the latest instalment of the triennial International Workshop on Cardiac Mechano-Electric Coupling and Arrhythmias (https://www.uniklinik-freiburg.de/experimental-cardiovascular-medicine/mec-2022.html). The issue starts with a tribute to our recently departed colleague, Dr Frederick Sachs, written by his long-time collaborator, Dr Tom Suchyna (Suchyna, 2024). Dr Sachs was a world-renowned biophysicist and a pioneer in the field of cardiac mechano-sensitivity. He was a frequent attendee and contributor to the Cardiac Mechano-Electric Coupling and Arrhythmias meeting and co-editor of the seminal textbook on the subject (Kohl et al., 2011). While Dr Sachs made extensive pathophysiological, conceptual, and technical contributions to the field, most notable was the discovery of mechano-sensitive ion channels in non-sensory cells (i.e., skeletal and cardiac muscle), first published in The Journal of Physiology in 1984 (Guharay & Sachs, 1984), and further characterised in a series of other papers in The Journal (Guharay & Sachs, 1985; Yang & Sachs, 1990) and elsewhere. Ultimately, Dr Sachs’ impact on our understanding of mechano-sensitive channels, cellular mechanics, and mechanical signalling, and their role in pathophysiology cannot be overstated. He will be dearly missed by those that new him well and those who did not have the pleasure but whose research has been influenced by his work. To get a sense of the love and respect that researchers in the field felt for Fred – as well as a sense of the man himself – please view some of the videos found at this website about his life and work: https://www.speaking-of-fred-sachs.org/colleagues-on-fred/. While the concept of cardiac mechano-electric coupling (MEC) has developed over the past 60 years from a collection of anecdotal reports met with widespread scepticism into an important consideration for physiological and pathophysiological heart function (Quinn et al., 2014), still much is unknown about the cellular and subcellular mechanisms involved (Cameron et al., 2023). The first series of papers in this Special Issue identifies novel subcellular mechanisms important for mechanically-induced functional responses in cardiomyocytes. One of the most recognised responses to cardiomyocyte stretch is the ‘Frank-Starling Law of the Heart’ (Katz, 2002), a prime example of ‘mechano-mechanical coupling’ (Quinn & Kohl, 2016) that leads to an immediate increase in contractile force and more efficient contraction (Han et al., 2022). In the study by Kaihara et al. (2023), they demonstrate that stretch-induced reactive oxygen species (ROS) production by NADPH oxidase 2 (NOX2) increases cardiomyocyte contractility. Experiments involved axial stretch of single cardiomyocytes isolated from wild-type and NOX2-deficient mice (NOX2−/−). In wild-type cardiomyocytes, stretch resulted in an increase in ROS production that was absent in the NOX2−/− animals. In cardiomyocytes from NOX2−/− mice, the lack of stretch-induced ROS was associated with lower contractility and a greater delay in the rise of contractile force and cytosolic Ca2+ concentration compared with wild-type cardiomyocytes. These findings were complemented by computational simulations that suggest augmented activation of ryanodine receptors by stretch-induced ROS causes an increase in the rate of sarcoplasmic reticulum Ca2+ release, resulting in a more rapid rise in intracellular Ca2+, which increases contractility to counteract an increase in preload. Interestingly, it has recently been shown by others in The Journal that similar effects appear to occur in cardiomyocytes to maintain contraction in response to increased afterload (Alim et al., 2022). Combined, these pre- and afterload-induced effects may represent an important means by which intercellular contractile heterogeneity is mediated across the billions of myocytes found in the heart. Another well-established response to stretch of spontaneously beating cardiomyocytes is the ‘Bainbridge Effect’ (Quinn & Kohl, 2022), which leads to a rapid increase in beating rate (MacDonald et al., 2020). In the study by Guo et al. (2023), they demonstrate that functional coupling between Piezo1 and transient receptor potential melastatin 4 (TRPM4) channels influences cardiomyocyte electrical activity. Experiments involved exposure of HL-1 mouse atrial cardiomyocyte-like cell cultures to a Piezo1 agonist as a surrogate for stretch-induced activation. Piezo1 activation resulted in an increase in beating rate, which was prevented by Piezo1 knockdown. The increase in beating rate was also reduced by inhibition or knockdown of TRPM4, suggesting functional coupling between the two channels. Interestingly, TRPM4 has also recently been shown by others in The Journal to be linked to a Ca2+-induced increase in cellular excitability and arrhythmias (Pironet et al., 2024), while Piezo1 has been shown to functionally couple to big K+ channels in human atrial fibroblasts in the context of atrial fibrillation (Jakob et al., 2021). Together, these results emphasise the importance of functional coupling between ion channels, adding further complexity to the molecular mechanisms underlying MEC. The composition, organisation and resulting mechanical stress of the plasma membrane is a crucial determinant of sarcolemmal mechano-sensitive ion channel function. In the study by Turner et al. (2023), they reveal a role of neutral sphingomyelinase in the loss of caveolae, increase in ceramide production and reduction in β-adrenergic response that occurs with chronic stretch of cardiomyocytes. Experiments involved subjecting engineered cardiac tissue composed of human induced pluripotent stem cell-derived cardiomyocytes and fibroblasts to a long-term cyclic stretch protocol with and without pharmacological inhibition of sphingomyelinase. Sphingomyelinase inhibition reduced the loss of caveolae-like structures and prevented the increase in ceramide production and the reduction in the inotropic response to β-adrenergic stimulation that occurred with chronic stretch. These findings were combined with blood lipidomic analysis from human subjects that showed an increase in circulating levels of ceramides in adults with hypertension, and mouse isolated heart experiments in which sphingomyelinase inhibition prevented pressure-induced left ventricular conduction slowing (which may relate to the conduction slowing caused by caveolae incorporation into the sarcolemma with acute stretch, as shown by others (Pfeiffer et al., 2014)). In the following study, Sommerfeld et al. (2024) demonstrate a sex-dependent risk of atrial arrhythmias in patients with desmosomal gene variants – which can lead to arrhythmogenic cardiomyopathy (Thiene et al., 2023). They show that plakoglobin deficiency (Plako+/−) causes pathological atrial electrical remodelling in young male hearts, which is exaggerated by androgenic anabolic steroids. Analysis of clinical records from a cohort of arrhythmogenic right ventricular cardiomyopathy (ARVC) patients revealed that ARVC is more common in males than females, and that patients with ARVC had an increased incidence of atrial arrhythmias. This analysis was followed by experiments in wild-type and Plako+/− male mice chronically exposed to an endogenous androgen steroid, which caused an increase in atrial expression of pro-hypertrophic, fibrotic, and inflammatory transcripts in both groups. In the Plako+/− mice, chronic androgen treatment caused adverse electrophysiological changes, including electrocardiogram P-wave abnormalities and reduced speed of atrial conduction, Na+ current, and action potential depolarisation rate and amplitude, associated with a decrease in NaV1.5 membrane clustering. The above papers, which focused on cardiomyocytes, are complemented by a study from Simón-Chica et al. (2024) – reviewed in the accompanying Journal Club article by Farahi and Pourteymour (Farahi & Pourteymour, 2024) – that reveals a stretch-induced membrane current in macrophages, the characteristics of which differ between bone marrow-derived and cardiac tissue-resident populations. Experiments involved cell-attached patch-clamp recordings and intracellular Ca2+ imaging in murine macrophages that were either differentiated in vitro from bone marrow or enzymatically isolated from cardiac tissue, complemented by mechano-sensitive ion channel gene expression data sourced from RNA sequencing of bone marrow and cardiac tissue-resident macrophages. Piezo1 and TRPV4 (which have recently been shown by others in The Journal to be involved in mechanically-induced responses in skeletal muscle afferents contributing to cardiac effects during exercise (Fukazawa et al., 2023)) were found to both be genetically expressed in bone marrow and cardiac macrophages. While functional membrane activity of TRPV4 (confirmed by Ca2+ entry during pharmacological activation) was also demonstrated in macrophages from both sources, Piezo1 activity was detected in the bone marrow-derived cells only. Importantly, these findings add additional importance to considering the diversity of cardiac cell populations when studying sources of mechanically-induced responses (Grandi et al., 2023; Kass, 2023). The next set of papers in this Special Issue focuses on physiological effects of mechanically-induced responses in cardiomyocytes, cardiac tissue and the whole heart. It is known that from the beginning of development, the heart's structure and electrical and mechanical function are dependent on its mechanical load (Baillie et al., 2023), which continues to evolve in the postnatal environment (Salameh et al., 2023). The study by Manfra et al. (2023) demonstrates that the maturation of t-tubules in cardiomyocytes during fetal development is dependent on the level of the heart's systolic load. Experiments involved examining t-tubule growth during fetal sheep development – in which there is a naturally occurring increase in systolic blood pressure – combined with its experimental manipulation. It was found that increasing load by plasma infusion or aortic occlusion accelerated t-tubule growth, while reducing load with an angiotensin-converting enzyme inhibitor did the opposite. Interestingly, changes in t-tubule density did not relate to changes in the formation of dyadic junctions (where L-type Ca2+ channels and the Na+-Ca2+ exchanger in t-tubules are in close proximity to ryanodine receptors in the sarcoplasmic reticulum) or the levels of their regulatory proteins (junctophilin-2, amphiphysin-2 [BIN1] and its partner proteins), indicating that distinct maturation signalling is involved. In another study from the same group, Ruud et al. (2024) demonstrate a compensatory increase in t-tubule density in response to a moderate (physiological) increase in cardiac preload or afterload in cardiomyocytes, but a decrease in density during pathological mechanical overload (which has recently been shown by others in The Journal to drive Ca2+-driven pro-arrhythmic behaviour (Heijman & Dobrev, 2023; MacLeod, 2023; Zhang, Ni, et al., 2023; Zhang, Smith, et al., 2023)). Experiments involved stepwise increases in the preload or afterload of working cultured rat papillary muscles, which revealed a bell-shaped relationship between load and cardiomyocyte t-tubule density. This was followed by in vivo mouse experiments, which showed that moderate increases in preload or afterload (1 week of aortic shunt or banding) resulted in t-tubule growth and compensated cardiac function. These results were complemented by analysis of cardiac tissue samples from human patients with moderately increased preload or afterload (due to mild mitral valve regurgitation or aortic stenosis), in which t-tubule proliferation was also observed. Enhanced t-tubule growth was associated with greater Ca2+ transient amplitude and upregulation of L-type Ca2+ channels, the Na+-Ca2+ exchanger, and regulators of t-tubule structure. In contrast, in mice and patients with a greater elevation of cardiac load, there was a decrease in t-tubule density. The next study, by Lewetag et al. (2023), investigated the hypothesis that long QT syndrome is an electro-mechanical disease, rather than purely electrical, showing that cardiac MEC plays a role in long QT-related arrhythmogenesis. It is now becoming well established that heterogeneous mechanical function in cardiac disease is a contributor to the induction and sustenance of ventricular tachyarrhythmias (Quinn, 2014). In this paper, experiments involved in vivo measurements of regional differences in electrical and mechanical activity of hearts in healthy rabbits and rabbits with drug-induced long QT syndrome (using a blocker of the rapid delayed rectifying potassium current, IKr) during increases in preload. Phase mapping cardiac MRI and a 24-lead electrocardiogram vest revealed that in the long QT syndrome rabbits, preload increased electrical repolarisation time and its spatial and temporal heterogeneity, compared with healthy animals. This effect of preload on repolarisation was confirmed ex vivo using voltage optical mapping of isolated beating hearts, which is being made possible by evolving technological imaging innovations recently reviewed in The Journal (Kappadan et al., 2023). In vivo, increased electrical heterogeneity was associated with reduced diastolic mechanical function and prolonged contraction, causing increased mechanical heterogeneity. Increased preload acutely prolonged repolarisation in both groups but was more pronounced in long QT syndrome rabbits, in which there was also an increase in spatial heterogeneity. The final collection of Special Issue papers comprises studies using in silico computational approaches to further our understanding of the mechanisms and effects of cardiac MEC. Computational modelling (Quinn & Kohl, 2013) and systems biology (Quinn & Kohl, 2011) are increasingly powerful tools for the analysis of experimental data and the integration of complex cardiac behaviour, building on decades of iteration between experiment and theory that has contributed to a more complete and cohesive understanding of cardiac pathophysiology. In the study by Van de Graaf et al. (2023), they benchmark the utility of available protein interaction databases – critical resources for network bioinformatics and integration of molecular experimental data (Chowkwale et al., 2023) – for use in logic-based biological network models of cardiac mechano-signalling, hypertrophy and fibrosis (which are known arrhythmogenic factors in diseases involving cardiac mechanical overload, such as hypertension (Khwaounjoo et al., 2022)). They found that while available databases successfully recovered central, well-conserved pathways, they were less successful at recovering pathways that are tissue-specific or relate to transcriptional regulation, highlighting cases in which manual curation is critical. This was followed by an illustration of their use for identifying novel signalling interactions, which revealed a role for Ca2+/calmodulin-dependent protein kinase II (CaMKII) phosphorylation of cAMP response element-binding protein and protein kinase C autophosphorylation in cardiomyocyte hypertrophy. The next study, by Mazhar et al. (2023), describes a novel computational electro-mechanical model of the human atrial cardiomyocyte, in which they included MEC effects. The new computational model was developed by coupling modified versions of existing human atrial cardiomyocyte action potential and contraction models, followed by calibration and validation against available human in vitro data. The model was employed to investigate mechanisms by which stretch contributes to a substrate for atrial fibrillation. Under conditions of post-operative atrial fibrillation (simulated by a reduction in sarco/endoplasmic reticulum Ca2+ ATPase activity), the model showed a decrease in the pacing frequency threshold for action potential and Ca2+ transient alternans, which was exacerbated by stretch in a magnitude-dependent manner due to decreased ryanodine receptor refractoriness and increased binding of Ca2+ by troponin C. In the following study by Buonocunto et al. (2023), they use computational simulations to systematically characterise stretch effects – through activation of cation non-selective, Ca2+-selective and K+-selective stretch-activated channels – on human ventricular electrophysiology. Simulations used a human action potential model calibrated with available experimental data from stretched isolated cardiomyocytes, including consideration of inter-species differences and disease-related stretch-activated channel remodelling, to analyse effects of varying stretch amplitude, timing and duration on the ventricular action potential. The simulations showed various effects, depending on stretch characteristics, including after-depolarisations, altered excitation threshold, triggered excitation, excitation suppression, and changes in action potential morphology, which were enhanced by disease-related stretch-activated channel remodelling. In the subsequent study, by Gerach and Loewe (2024), they used a four-chamber, electro-mechanical whole heart model to investigate the effects of multiple known MEC mechanisms; namely, the influence of myocardial deformation on electrical conduction (modelled by its inclusion in the tissue diffusion tensor), stretch-activated channel activity, and enhanced binding of Ca2+ by troponin C on cardiac function. It was found that inclusion of deformation in the tissue diffusion tensor primarily affected excitation propagation, slowing conduction and heterogeneously increasing the time of myocardial activation, while stretch-activated channels and troponin Ca2+ binding heterogeneously slowed repolarisation. Most notably, the resulting longer action potentials and steeper repolarisation gradients in the atria led to after-depolarisations. In addition, troponin Ca2+ binding affected the heart's mechanical activity, increasing developed pressure and stroke volume. In the last study in this series, by Sung et al. (Sung et al., 2023) – which is reviewed in the accompanying Journal Club article by Negin Amini (Amini, 2024) – the authors demonstrate the utility of automated, deep learning-based reconstruction of personalised computational electrophysiological heart models to guide patient-specific treatment of ventricular tachycardia. This involved development of a deep convolutional neural network for segmentation of the left ventricle from contrast-enhanced computed tomography images, which was applied to images from a cohort of patients with accompanying expert manual segmentations, followed by simulations of rapid pacing to induce ventricular tachycardias. It was found that in simulations with heart models generated from automated and manual segmentations there was a similar propagation of excitation, along with a similar arrhythmia inducibility, a strongly correlated number of induced ventricular tachycardias, and co-localised re-entrant circuits. Taken together, the results demonstrate the potential utility of automated computational model generation for patient-specific arrhythmia risk stratification and ablation targeting. The final paper in this Special Issue, by Pandit et al. (2024), elucidates a direct clinical implication of cardiac MEC by reviewing the evidence and potential underlying mechanisms for its role in the recurrence of ventricular fibrillation in out-of-hospital cardiac arrest. While ventricular fibrillation termination with defibrillation in out-of-hospital cardiac arrest patients has an impressively high success rate, a persistent challenge is the high rate of ventricular fibrillation recurrence during post-shock cardiopulmonary resuscitation, whose mechanisms are et al., 2021). the authors a for a role of MEC in ventricular fibrillation occurring during in the out-of-hospital by the clinical evidence of fibrillation recurrence and examining the potential of heart level MEC and pathways from clinical and experimental that the papers in this Special Issue of The Journal of Physiology to Cardiac Mechano-Electric Crosstalk a greater understanding and for the critical role of cardiac MEC at various spatial and temporal for the development, and of the heart's electrical and mechanical activity, and technological innovations into its importance in and The is not for the or of by the than be to the for the and the paper and to be for all aspects of the in that to of its or are investigated and as authors for and all those who for are The is the Cardiac to the of and to the and of to the of in to and the and of to is an of the