Mechano‐chemo‐transduction in cardiac myocytes

Abstract

The heart has the ability to adjust to changing mechanical loads. The Frank–Starling law and the Anrep effect describe exquisite intrinsic mechanisms the heart has for autoregulating the force of contraction to maintain cardiac output under changes of preload and afterload. Although these mechanisms have been known for more than a century, their cellular and molecular underpinnings are still debated. How does the cardiac myocyte sense changes in preload or afterload? How does the myocyte adjust its response to compensate for such changes? In cardiac myocytes Ca²⁺ is a crucial regulator of contractile force and in this review we compare and contrast recent studies from different labs that address these two important questions. The ‘dimensionality’ of the mechanical milieu under which experiments are carried out provide important clues to the location of the mechanosensors and the kinds of mechanical forces they can sense and respond to. As a first approximation, sensors inside the myocyte appear to modulate reactive oxygen species while sensors on the cell surface appear to also modulate nitric oxide signalling; both signalling pathways affect Ca²⁺ handling. Undoubtedly, further studies will add layers to this simplified picture. Clarifying the intimate links from cellular mechanics to reactive oxygen species and nitric oxide signalling and to Ca²⁺ handling will deepen our understanding of the Frank–Starling law and the Anrep effect, and also provide a unified view on how arrhythmias may arise in seemingly disparate diseases that have in common altered myocyte mechanics.

Abbreviations

CaMKII

Ca2+–calmodulin‐dependent protein kinase II

NOS

nitric oxide synthase

eNOS

endothelial NOS

MCT

mechano‐chemo‐transduction

nNOS

neuronal NOS

NO

nitric oxide

NOS

nitric oxide synthase

NOX2

nicotinamide adenine dinucleotide phosphate oxidase 2

ROS

reactive oxygen species

X‐ROS

NOX2‐derived ROS

Knowledge gap in mechano‐chemo‐transduction

In every heartbeat, cardiac myocytes generate mechanical force to pump blood into circulation to meet the body's demands for oxygen and fuel. The heart must continuously adjust the force of contraction to compensate for changes in mechanical load resulting from changes in posture, physical activity and emotional state. The Frank–Starling law and the Anrep effect describe the intrinsic auto‐regulatory mechanisms that enable the heart to respond and adapt to changes in preload and afterload to maintain adequate cardiac output. Changes in preload or afterload must, by Laplace's law, change the wall stress borne by the cardiac myocytes. The Frank–Starling law and the Anrep effect were discovered more than a century ago yet their cellular and molecular bases remain unclear despite intensive study (de Tombe et al. 2010; Bollensdorff et al. 2011; Cingolani et al. 2013). A critical knowledge gap is how the myocyte senses mechanical load and transduces the load information to biochemical reactions to affect cell function, a process called mechano‐chemo‐transduction (MCT). Deciphering the key molecules and the signalling pathways in MCT is necessary for understanding how the heart auto‐regulates contractility under physiological loading, but also for understanding how arrhythmias and perhaps cardiomyopathies develop under excessive pathological loading (Knoll et al. 2002).

During contraction, myocytes in situ experience longitudinal shortening, transverse thickening, and shearing strains (Waldman et al. 1985). The in situ mechanical environment of cardiac myocytes is complex because regional differences exist in anatomical geometry and cell orientations, and there is considerable mechanical interaction between cells (Waldman et al. 1985; Young et al. 2001). The mechanical environment of isolated single cardiac myocytes is simpler and can be experimentally controlled to allow detailed study of the molecular mechanisms underlying MCT. Here we review the recent work in developing new experimental tools to capture the acute MCT process in action during beat‐to‐beat myocyte contraction. Because of space constraints, it is impossible to reference all relevant original articles in the field. We will focus on addressing how the ‘dimensionality’ of the three‐dimensional mechanical milieu of the adult cardiac myocyte reveals different modes of mechanosensing and mechanotransduction.

Classes of mechanosensors and their cellular locations

Two different classes of mechanosensors for cardiac myocytes are readily conceivable. One class, based on changing the area and thickness of the lipid bilayer that make up the sarcolemma when the myocyte is stretched, involves changes in surface charge density and specific capacitance that would change the electric field (from Gauss's law) and the transmembrane voltage (Shapiro et al. 2012). The other class comprises proteins (ion channels, enzymes, etc.) that change activity (conductance, enzymatic activity, etc.) resulting from conformational changes induced by mechanical forces imposed on the myocyte. In this paper we focus on this latter class of mechanosensors. We find it conceptually useful to think of the mechanosensors as molecular ‘springs’ that are either wholly inside the myocyte (S₁ and S₂ in Fig. 1) or having a part on the surface (S₃ and S₄). We refer to S₁ and S₂ as internal mechanosensors and S₃ and S₄ as surface mechanosensors. For clarity S₁ and S₂ are drawn as separate sensors to detect orthogonal stresses, but a single sensor arranged at some intermediate angle can respond to both directions of stress. As the myocyte is stretched or contracts by a fraction, ε_m, the sensor (we assume) experiences its own strain, ε_s. Strain is defined by change in length divided by original length, ε = ΔL/L ₀ (reference cell and mechanosensors shown are in the left panel of Fig. 1). The negative myocyte strain, ε_m, during contraction is often called the fractional shortening. Fractional shortening at the whole cell or sarcomere level is routinely measured in the lab but sensor strain, ε_s, at the molecular level is admittedly a hypothetical – but physically plausible – construct.

Figure 1. Location of mechanosensors and how they change during contraction.

Internal mechanosensors S₁ and S₂ respond to orthogonal strains. Surface sensor S₃ lies on the neutral plane. ε_m is the myocyte strain (fractional shortening) and ε_s is the sensor strain. Left panel depicts a myocyte embedded in‐gel at resting state. Right panel depicts a myocyte in‐gel at contracted state when the internal and surface mechanosensors are under mechanical loading.

Cell strain and sensor strain

To approximate the three‐dimensional (3‐D) mechanical environment, we developed the cell‐in‐gel system that embeds single myocytes in a 3‐D viscoelastic hydrogel (Jian et al. 2014). The contracting myocyte does work on the gel because the boronate‐crosslinked polyvinyl alcohol gel matrix binds to cell surface glycans (Luo et al. 2009). Mechanical analysis shows that the cell experiences longitudinal tension (along the cell's long axis), transverse compression (along the short axis) and surface traction forces (vector sum of the shear and normal forces) in this system (Shaw et al. 2013). A contracting myocyte must pull against the gel in the longitudinal direction and, because of the isovolumic constraint, must push against the gel in the transverse direction. In this gel environment, contraction or stretching of the myocyte (by ε_m) is expected to cause strains in both internal and surface mechanosensors (S₁–S₄ in Fig. 1). S₃ and S₄ bind to the gel and so experience strain when the cell contracts or stretches. However, S₃ and S₄ can respond differently because of their different longitudinal positions on the cell. During contraction both S₃ and S₄ are compressed but only S₄ experiences shearing (the end attached to the myocyte moves to the left). S₃ on the cell's neutral plane (located at a point along the long axis, near the centre, where no displacement occurs during contraction or stretch) experiences no shear. We do not think that surface mechanosensors primarily sense shear for two reasons. First, a straightforward geometrical argument shows that the sensor strain due to shear only (no transverse compression or elongation) is proportional to the square of the myocyte strain ε_s ∝ ε_m ². Thus for typical fractional shortening of ∼10%, the sensor strain would be only ∼1%. This low sensitivity would make sensors depending on shear unlikely to be effective load sensors. Second, because surface displacement (not strain) varies linearly from zero at the neutral plane to maximal at the ends of the cell, there would be a gradient of sensor strain. Assuming that sensor strain translates to some effect on Ca²⁺ (see below) one would then expect a Ca²⁺ gradient along the longitudinal axis. We observed no such gradient (Jian et al. 2014), so we think that – at least in the cell‐in‐gel system – either the surface mechanosensors are not responding primarily to shear or the stress from sensors is somehow integrated and evenly distributed throughout the cell perhaps because of the tensegrity structure of myocyte (White, 2011) .

On the other hand, when the myocyte contracts or is stretched the cell surface moves transversely uniformly so the surface mechanosensors are compressed or stretched to the same extent ε_s everywhere along the longitudinal axis. Furthermore, in this case, sensor strain is proportional to myocyte strain, ε_s ∝ ε_m. The high sensitivity and longitudinal uniformity of response make us favour the interpretation that surface mechanosensors are compressed or stretched in the gel or in situ.

Our working hypothesis is that the magnitude of the sensor strain is positively related to the magnitude of the intracellular signal: bigger sensor strain, bigger signal. This hypothesis is saved from being a mere platitude because it dissociates sensor strain from cell strain. Because of this dissociation, equal cell strains need not produce the same sensor strains in different experimental systems. This difference can be exploited to distinguish surface and internal mechanosensors.

Interrogating different mechano‐chemo‐transduction pathways

The relaxed myocyte is depicted in the top left panel (Load‐free) of Fig. 2. In this reference state neither the cell nor the mechanosensors are strained. The perfect circles represent this strain‐free state of the mechanosensors. Different experimental systems will cause different strains on the mechanosensors. Physiological saline solution offers little mechanical resistance (load‐free) to the contracting myocyte (Fig. 2, Load‐free, right panel) so the surface mechanosensors experience little or no strain while the internal mechanosensors are strained (note the change from circles to ellipses).

Figure 2. How different experimental systems distinguish between surface and internal mechanosensors.

Surface (yellow) and internal (red and green) mechanosensors are depicted as springs. As the myocyte is stretched (left column) or contracts (right column) surface mechanosensors experience different strains (change in length divided by original length) depending on experimental system. The ellipticity of the mechanosensor represents the degree of strain. A perfect circle indicates zero strain. In the 1‐D system, the blue cylinders at the cell edges are the glass rods or carbon fibres. In the 2‐D system the blue sheet represents the stretchable membrane. In the 3‐D system, the blue cylinder represents the hydrogel.

3‐D cell‐in‐gel system

The polyvinyl alcohol matrix in the 3‐D cell‐in‐gel system can be tuned to have Young's modulus between 1 and 20 kPa and viscosity of about 1 Pa s (at a strain rate of 1 s^–1). These values are comparable to myocardial stiffness and viscosities measured in the human heart (Pislaru et al. 2014). Figure 2 (3‐D) depicts the myocyte in the 3‐D gel (blue cylinder). Because of the gel's viscoelasticity the cell does work against the gel during contraction (Shaw et al. 2013) and presumably part of the work goes into compressing and shearing the surface mechanosensors (right panel). Note the variation in shearing along the long axis of the cell. The mechanosensor at the centre of the cell (the neutral plane) undergoes no shear while those at the edge undergo the most shear. The internal mechanosensors experience the same kinds of strain (but perhaps of different magnitude) as in load‐free conditions. Experimental evidence from our lab and others suggests that compression of the surface mechanosensors activates nitric oxide synthase (NOS).

Myocytes contracting in‐gel showed an increased magnitude of Ca²⁺ transient in systole, compared with load‐free cell contraction in bathing solution (Jian et al. 2014). This increase of the Ca²⁺ transient enhances contractility to compensate for the greater mechanical load, possibly contributing to the Anrep effect. However, the MCT also caused spontaneous Ca²⁺ sparks during diastole that sometimes arise synchronously to form Ca²⁺ waves (Awasthi et al. 2015), which could trigger afterdepolarizations and premature action potentials that provide substrate for cardiac arrhythmias (Kass & Tsien, 1982; Berlin et al. 1989; Katra & Laurita, 2005). MCT in the cell‐in‐gel system involves several important signalling pathways. Both of the constitutive NOS isoforms – neuronal NOS (nNOS or NOS1) and endothelial NOS (eNOS or NOS3) – participate in elevating the systolic Ca²⁺ transient; however, only nNOS plays a critical role in the induction of diastolic Ca²⁺ sparks and waves. We also found that nicotinamide adenine dinucleotide phosphate oxidase 2 (NOX2), whose importance in mechanosensing was identified by Prosser et al. (2011), is also involved in generating Ca²⁺ sparks and waves. We cannot rule out the possibility that surface mechanosensors may also activate reactive oxygen species (ROS) production, since the surface sensors have not been studied in isolation. Furthermore, downstream from NOX2 and NOS signalling, Ca²⁺–calmodulin‐dependent protein kinase II (CaMKII) was found to mediate the MCT effect on Ca²⁺ handling (Jian et al. 2014).

Comparison with 1‐D systems

An important experimental method for studying MCT uses a pair of carbon fibres or glass rods attached to the myocyte to stretch the cell along the long axis (Le Guennec et al. 1990; Alvarez et al. 1999; Calaghan & White, 2004; Iribe et al. 2009; Prosser et al. 2011). We refer to this system, shown in Fig. 2 as a 1‐D stretching system because mechanical loading on the cell is directed primarily along one dimension. A critical difference between the 3‐D cell‐in‐gel system and the 1‐D system is the mechanical environment of the myocyte. In the cell‐in‐gel system the myocyte pulls or pushes against the gel during stretch or contraction so the surface mechanosensors may be stretched or compressed. In contrast, in a 1‐D system the myocyte is bathed in physiological saline solution, which is not elastic and has very low viscosity compared to the gel (1 Pa s vs. ∼10⁻³ Pa s, the viscosity of water). Therefore, the surface mechanosensors are unlikely to experience significant strain (small ε_s) even though myocyte strain (ε_m, fractional shortening) might be large. This dissociation between myocyte strain and surface mechanosensor strain is illustrated in Fig. 2 by the surface mechanosensors remaining circular despite stretching or contraction in the 1‐D system.

Using the 1‐D system, Iribe et al. (2009) found that stretching the myocyte at resting state induced spontaneous Ca²⁺ sparks. Prosser et al. further discovered that the MCT is mediated by NOX2‐derived ROS (X‐ROS) signalling (Prosser et al. 2011), and cyclic stretching of the resting myocyte elevates the ROS level (Prosser et al. 2013). Importantly, the non‐specific NOS inhibitor N ^ω‐nitro‐l‐arginine methyl ester (l‐NAME) had no effect on the stretch‐induced Ca²⁺ sparks (Iribe et al. 2009). By contrast, l‐NAME abolished the Ca²⁺ sparks in the cell‐in‐gel system (Jian et al. 2014). On the other hand, we also found that X‐ROS signalling is present in myocytes contracting in gel. The differences and similarities between the 1‐D stretching system versus the 3‐D cell‐in‐gel system point to distinct signalling pathways activated by internal and surface mechanosensors. Internal mechanosensors are presumably activated during stretching or contraction in both 1‐D and 3‐D systems but the surface mechanosensors are likely to be activated only in the 3‐D system. Therefore, these data suggest that X‐ROS signalling is activated by internal mechanosensors while NO signalling is activated by surface mechanosensors.

The idea that surface mechanosensors activate NOS is further supported by the work of Petroff et al. (2001), who embedded myocytes in agarose and then encapsulated the cell and agarose in a deformable tube. In this experimental setting, because agarose is inelastically deformable and does not bind to biomolecules to any significant extent (Freifelder, 1982), stretching the tube squeezes the encapsulated cells to exert compressive stress on the cell and presumably the surface mechanosensors as illustrated in Fig. 2 (3‐D system, left). Petroff et al. (2001) found that stretching the agarose tube caused spontaneous Ca²⁺ sparks in resting myocytes, and also increased Ca²⁺ transient in electrically stimulated myocytes, and this mechanotransduction is mediated by eNOS‐dependent NO signalling.

Comparison with 2‐D systems

Investigation using a 2D‐stretching system provides further evidence to support the idea that surface mechanosensors work via NOS. Garbincius & Michele (2015) cultured adult mouse ventricular myocytes (treated with blebbistatin to impair contraction) on laminin‐coated membrane, and then stretched the membrane to apply shear stress on cells (Fig. 2, 2‐D). After cyclically stretching the membrane at 1 Hz for 1 h, they measured the intracellular NO level, and found that shear stress increased nNOS‐derived NO. An important difference between the 2‐D and 3‐D systems is that in the former, the surface mechanosensors not facing the membrane are exposed to bath solution and experience very little strain. Thus in the 2‐D system, mechanotransduction signals come from internal mechanosensors and only a fraction of the surface mechanosensors, while in the 3‐D system all surface mechanosensors contribute to signalling.

As described above, different experimental systems can be used to interrogate different mechanosensors and MCT pathways. The 1‐D systems can be used to interrogate the internal mechanosensors separate from the surface mechanosensors; it is designed for studying adult ventricular myocytes (rod‐shaped cells). The 2‐D systems can be used to interrogate the surface mechanosensors, but it is difficult to attach adult cardiac myocytes on a 2‐D substrate. The 2‐D systems are often used for cultured cells such as neonatal ventricular myocytes that can attach to the deformable membrane. The advantages of the 3‐D cell‐in‐gel system are that the gel is easy to use, can embed freshly isolated adult cardiac myocytes and any cell type, and resembles more closely the 3‐D in situ mechanical environment. Because many heart diseases occur in adult hearts and especially in ageing hearts, it is important to control mechanical load on adult cardiac myocytes in studying mechanotransduction and heart disease mechanisms.

Mapping MCT pathways: internal mechanosensors and surface mechanosensors

What are the mechanosensors that sense different types of mechanical stresses? The mechanical structure of striated myocytes is composed of a sarcomeric cytoskeleton network with a remarkable axial and transverse order of filaments that form crystalline‐like lattices (Gautel & Djinovic‐Carugo, 2016). This intracellular mechanical network is also linked through the transmembrane costamere structure to the extracellular matrix (Ervasti, 2003). During the cardiac cycle, contraction and stretching of myocytes exert mechanical stresses on all these mechanical scaffolds. Based on the location of molecules in the myocyte's architecture, it is plausible that surface traction including normal and shear stresses can be sensed by surface mechanosensors, and tension and compression can act on internal mechanosensors and load‐bearing molecules in the sarcomeric cytoskeleton network.

For example, the giant protein titin spans half of the sarcomere to connect z‐disk, I‐band, A‐band and M‐line structures (Labeit et al. 1990; Bang et al. 2001). When stretched during diastolic filling, the extensible I‐band segment of titin lengthens and develops passive tension (Granzier & Irving, 1995). When relaxed after contraction, the Ig domains of titin recoil and provide restoring forces (Helmes et al. 1996; Preetha et al. 2005). Thus titin can sense both tension and compression during myocyte lengthening and contraction in the cardiac cycle. The passive force of titin is dependent on its compliance, which is regulated by phosphorylation of the N2B spring element by PKA (Yamasaki et al. 2002) and CaMKIIδ (Hidalgo et al. 2013; Perkin et al. 2015). Because the CaMKIIδ activity is upregulated by MCT in the cell‐in‐gel system (Jian et al. 2014), mechanical loading is expected to reduce myocyte stiffening due to passive force generation by titin. Furthermore, titin mutations are the most common cause of dilated cardiomyopathy (Gigli et al. 2016), highlighting the essential role of titin in tension‐bearing and mechanotransduction.

Another important load‐bearing molecule in the cytoskeleton network is the microtubule, which provides a compression‐resisting structure in the myocyte. It was shown that microtubules affect the longitudinal shear stiffness, and this is consistent with microtubules serving as compressive resistance struts within a tensegrity model (White, 2011). Khairallah et al. (2012) showed that X‐ROS amplified Ca²⁺ influx through stretch‐activated channels in skeletal muscle of mdx mice (a model Duchenne muscular dystrophy). They also showed that MCT by X‐ROS signalling depends on microtubules. Disrupting the dense microtubular network in adult mdx mice resulted in little ROS production and Ca²⁺ influx. A recent seminal study by Robison et al. (2016) showed that detyrosinated microtubules interact with sarcomere to form a load‐bearing structure parallel to the myofilament. The microtubules buckle during myocyte contraction to resist compression; increased detyrosination of microtubules enhances their binding to the sarcomere and stiffen the myocyte in some forms of heart disease. Disrupting microtubules was found to impair the MCT through the NOX2–ROS pathway in the 1‐D system (Prosser et al. 2011), again demonstrating a significant role of microtubules in MCT.

While stretch and compression can be sensed by internal mechanosensors, surface traction and compression can be sensed by cell‐surface mechanosensors. As an illustrative example, dystroglycans outwardly link to the extracellular matrix and inwardly link to dystrophin to form the dystrophin glycoprotein complex (DGC). The DGC, together with the integrin–talin–vinculin complex, form the costamere structure that links myofibrils and Z‐disks to the sarcolemma and forms transmembrane linkages to the extracellular matrix (Anastasi et al. 2009). Costameres serve as radiating beams in the mechanical scaffold to provide structural support during beat‐to‐beat contraction (Danowski et al. 1992). In muscular dystrophy, mutations in the DGC lead to skeletal and cardiac muscle dysfunction, arrhythmias and sudden death (Judge et al. 2011; Groh, 2012; Fayssoil et al. 2013).

The link DGC to NOS has been well established in skeletal and cardiac muscle. In cardiomyocytes, nNOS is found to bind to α1‐syntrophin in the DGC (Williams et al. 2006) and to dystrophin (Lai et al. 2013). Studies using a dystrophin‐null mouse model (mdx) found that the mdx hearts have decreased nNOS, unchanged eNOS, and increased iNOS expression (Bia et al. 1999), and are extremely vulnerable to pressure overload (Kamogawa et al. 2001). Expression of the nNOS transgene could prevent the progressive ventricular fibrosis and greatly reduced myocarditis of mdx mice (Wehling‐Henricks et al. 2005). Mechanistic study in the 2‐D stretching system (using a membrane coated with laminin that binds to dystroglycans and exerts shear stress during stretching) shows that cyclic stretch increased nNOS‐derived NO in ventricular myocytes; in contrast, the stretch‐activated NO synthesis was impaired in the mdx model (Garbincius & Michele, 2015). Most recently, Casadei and colleagues found the dystrophin–nNOS link in atrial myocytes. In their study, microRNA‐31 repression of dystrophin translation caused nNOS translocation and depletion in human and goat atrial fibrillation (Reilly et al. 2016).

In the cell‐in‐gel system, because the cell‐surface glycans are tethered to the gel by boronate–polyethylene glycol crosslinking (Luo et al. 2009), surface traction imposes mechanical stress on glycosylated molecules including dystroglycans. Indeed, downstream in the dystrophin glycoprotein complex, dystrophin‐associated nNOS was found critical for mediating MCT induction of Ca²⁺ sparks in ventricular myocytes (Jian et al. 2014).

So far our focus has been on the MCT pathways involving NOS and NOX2 signalling to the Ca²⁺ handling system, but any conformational change in a protein (channel, enzyme, transporter, etc.) could potentially alter its activity. Thus, in principle, many proteins could sense mechanical force. For example, some ion channels such as stretch‐activated channels (Guharay & Sachs, 1984; Calaghan & White, 2004), the sodium channel Nav1.5 (Morris & Juranka, 2007), the Na⁺/H⁺ exchanger (Calaghan & White, 2004; Alvarez et al. 1999), the canonical transient receptor potential (TRPC6) channel (Seo et al. 2014) and the TRP vanilloid type 2 (TRPV2) channel (Lorin et al. 2015) are expressed in heart and are found to be mechanosensitive. Their localizations in different subcellular domains may position them to sense different mechanical stresses. Nav1.5 is on the intercalated disk and surface membrane but absent from transverse tubules (t‐tubules) (Westenbroek et al. 2013); TRPC6 (Dyachenko et al. 2009), TRPV2 (Lorin et al. 2015) and the Na⁺/H⁺ exchanger (Yang et al. 2002) are in the t‐tubules but not on the surface sarcolemma. However, it is still far from clear how the menagerie of mechanosensitive molecules is integrated so that the heart can respond to changes in mechanical loads. Further studies can be facilitated by development of innovative techniques to control mechanical stresses on the myocytes.

A mechano‐centric view of arrhythmias

Mechanical stresses are associated with various heart diseases. For example, hypertension, muscular dystrophy and dilated cardiomyopathy are seemingly disparate diseases but they share two common features. First, the risk of arrhythmias is high in these diseases (hypertension: McLenachan et al. 1987; Hennersdorf & Strauer, 2001; Saadeh & Jones, 2001; dilated cardiomyopathy: Meinertz et al. 1984; Neri et al. 1986; muscular dystrophies: McNally & MacLeod, 2005; Groh, 2012). Second, all these diseases involve altered myocyte mechanics. Hypertension exerts chronic pressure overload on ventricular wall. Muscular dystrophies involve mutations in the MCT machinery. Dilated cardiomyopathy involves high ventricular wall tension because of thinning of the wall and increased chamber radius. The high wall tension demands that myocytes contract more strongly to maintain a given stroke volume.

We described MCT mechanisms by which myocytes could sense mechanical load and alter Ca²⁺ dynamics to compensate for loading. The MCT pathways provide autoregulation of Ca²⁺ signalling and contractility in normal hearts, but may cause Ca²⁺ dysregulation in diseased hearts. As discussed, excessive mechanical loading can cause a high rate of diastolic Ca²⁺ sparks, which could trigger spontaneous Ca²⁺ waves (Cheng et al. 1993, 1996; Izu et al. 2001, 2006; Thul et al. 2012) leading to arrhythmogenic activities (Kass & Tsien, 1982; Berlin et al. 1989; Katra & Laurita, 2005). Interestingly, hypertensive patients also experience greater variability in mean arterial pressure than normotensives (Mancia et al. 2003). The many incidences of elevated blood pressure fluctuations on an already increased tension provide many chances for high frequency spontaneous Ca²⁺ sparks and waves to occur that could lead to arrhythmias. Like the lottery, the more times you play the more chances to lose. In the same vein, muscular dystrophy mutations could disrupt MCT pathways to cause Ca²⁺ dysregulation and arrhythmogenic activities. Various mutations also present opportunities to understand what happens when parts of the MCT pathways become compromised. Dilated cardiomyopathy inevitably involves high wall tension caused by dilation of the ventricular chamber and thinning of the wall, which is also expected to cause high frequency of Ca²⁺ sparks and waves that provide substrate to cardiac arrhythmias.

A mechano‐centric view of arrhythmias, although perhaps parochial, highlights the significant impact of mechanical stresses on setting up conditions for arrhythmias by MCT effects on Ca²⁺ handling. This viewpoint suggests that novel antiarrhythmic drugs could target molecules in the MCT pathways such as nNOS or X‐ROS to correct the Ca²⁺ dysregulation engendered by excessive mechanical loading. Understanding the MCT pathways and the key molecular determinants may hold the key to developing drug therapies to reduce arrhythmias and cardiomyopathy in patients with hypertension, muscular dystrophy and dilated cardiomyopathy.

Additional information

Author contribution

Y.C.‐I.: Conception and design; financial support; administrative support; provision of study materials or patients; collection and assembly of data; data analysis and interpretation; manuscript writing. L.I.: conception and design; financial support; administrative support; provision of study materials or patients; collection and assembly of data; data analysis and interpretation; manuscript writing. Both authors have approved the final version of the manuscript and agree to be accountable for all aspects of the work. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.

Funding

This work was supported by grants from the National Institutes of Health (NIH) R01HL90880 to L.T.I. and Y.C.‐I., American Heart Association 14GRNT20510041 to Y.C.‐I., and NIH R01HL123526 to Y.C.‐I. and L.T.I.

Biography

Ye Chen‐Izu and Leighton T. Izu collaborate in an interdisciplinary team to combine experiments with mathematical modeling for in‐depth analyses of the interaction of three dynamics systems – electrical, Ca²⁺ signaling, and mechanical systems – that govern cardiac function and heart diseases.