Cardiac fibroblasts can respond to mechanical deformation of the plasma membrane with characteristic changes of their membrane potential. These changes of fibroblasts membrane potential are determined by operation of stretch-inactivated channels. These channels, mainly permeable for sodium ions, are activated by compression of the cell leading to depolarization, and are inactivated by stretch, which in turn leads to hyperpolarization. Thus, cardiac fibroblasts function as mechanoelectric transducers in the heart and represent the cellular substrate for a cardiac mechanoelectrical feedback mechanism. Increased sensitivity of the cardiac fibroblasts to mechanical changes contributes to electrical instability and arrhythmia after myocardial infarction. Recent findings indicate that these processes involve the transfer of electrical signals via gap junctions. In this article we will discuss the recent progress in the electrophysiology of cardiac fibroblasts and their role in mechanoelectric feedback in healthy and diseased hearts.

Introduction

In addition to cardiomyocytes, fibroblasts in the heart have been reported recently to function as mechano-electric transducers and provide a potential source for arrhythmia. Fibroblasts and the extracellular matrix, which is produced by these cells, comprise a volume fraction of approximately 5–10% of the cardiac tissue [2]. Fibroblasts constitute the most numerous non-myocyte cell population (more than 90% of the non-myocyte cells) in the heart [13]. Other cells, including vascular smooth muscle and endothelial cells represent smaller fractions of this population [1]. In the atrial sinus node region, fibroblasts and the connective tissue occupy between 45% [64] and 75% [7] of the volume. All investigated sinoatrial nodes (in rabbit, guinea pig, cat, and pig) contained large amounts (45% or more) of collagen [53]. Recent studies [4] on the rabbit sinoatrial node region, revealed two spatially distinct fibroblast populations, which expressed different types of connexins, the gap junction forming proteins. Fibroblasts are coupled via connexin40 (Cx40) in fibroblast-rich areas devoid of myocytes, and by Cx45 in regions of the node where fibroblasts intermingle with myocytes. In contrast to the cardiomyocytes, fibroblasts are electrically non-excitable cells, and their electrophysiological function remained unexplored until recently. Novel studies demonstrated the existence of mechanosensitive ion channels in the plasma membrane of cardiac fibroblasts [21, 23]. These channels were similar to the mechanosensitive ion channels in other cell types (for review see reference [61]), suggesting that cardiac fibroblasts can transform mechanical stimuli into electrical signals [21, 23], which has been postulated in earlier studies [39]. If one assumes a role for cardiac fibroblasts in the mechano-electric feedback mechanism of the heart, the question arises as to how membrane potential changes that occur in cardiac fibroblasts upon mechanical stimulation, are transmitted to the adjacent cardiomyocytes. Over decades, fibroblasts in the heart were thought to be electrically isolated from the cardiomyocytes. However, novel results testify that fibroblasts form a coupled network of cells, which is structurally and functionally interconnected to the myocytes [3, 4, 40]. Thus, the formation of single gap junction channels between cardiac myocytes and fibroblasts may facilitate the transfer of electrical signals between the two cell types [31, 48].

In this article we will review some of the recent advances in the electrophysiology of cardiac fibroblasts. We will also discuss the role of cardiac fibroblasts in the mechano-electric feedback in healthy and diseased hearts.

Electrical properties of cardiac fibroblasts

The electrical properties of cardiac fibroblasts were studied for the first time in 1978 [24, 36]. The intracellular recordings were accomplished on multicellular tissue preparations with the use of fine microelectrodes [26, 28, 29, 30, 32, 34, 39, 40, 41]. Fibroblasts in rat atrial tissue had a resting membrane potential (E_m) of approximately -22 mV and input resistances of ≈0.5 GΩ [29, 34]. A typical feature of the cells was the sensitivity of their E_m to mechanical deformation of the plasma membrane, i.e. during the spontaneous atrial contractions.

More recently, fibroblasts were isolated enzymatically from atria and ventricles. The membrane currents of these freshly isolated cells were studied by means of the patch-clamp technique [21, 23]. Examples, typical for the mechano-sensitive membrane currents of cardiac fibroblasts are presented in Figs. 1 and 2. In the absence of electrical pulses, the membrane was clamped to a holding potential of -45 mV, a value slightly more negative than the normal resting potential of the fibroblasts. From this holding potential, application of test pulses with 140 ms duration adjusted the membrane potential to values between -100 and +50 mV.

Figure 1

Series of whole-cell currents from cardiac fibroblasts in the absence (A) and presence (B) of stretch (2 μm). (C) Effect of 8 μM Gd³⁺ on the whole-cell currents during sustained stretch Changes of the holding current are marked by arrows. (more...)

Figure 2

Whole-cell currents from cardiac fibroblasts in the absence (A) and presence (B) of mechanical compression (2 μm). (C) Effect of 8 μM Gd³⁺ on the whole-cell currents during sustained compression. (From Kamkin et al. [21], with permission (more...)

Compression

Fig. 2 shows how net membrane currents are modulated by compression. Mechanical compression of the fibroblasts shifted the holding current at -45 mV to more negative values (compare the initial traces in Fig. 2B and Fig. 2A). Compression increased the current amplitudes during the depolarizing clamp steps without changing their time course. At negative potentials, the currents were more negative than under control conditions without compression. Hence, mechanical compression increased the membrane conductance of the atrial fibroblasts. Finally, when 8 μM Gd³⁺ was added to the bath solution, not only the compression-induced currents but also a large portion of the currents in the absence of mechanical compression was blocked (Fig. 2C). These findings strongly suggest that mechanical compression activates a mechanosensitive conductance in the cardiac fibroblasts.

In addition to MSC, cardiac fibroblasts were reported express voltage-dependent potassium K_v channels [21, 23]. Since Gd³⁺ was found to reduce also K⁺ currents, a detailed analysis of these potassium currents and their potential mechanosensitivity has not yet been accomplished.

The currents at the end of the 140 ms long voltage-pulses were assembled in current-voltage relations (I-V curves, Fig. 3). Without mechanical stimulation, the I-V curves intersected the voltage axis at -37 mV. This zero-current potential corresponds to the normal resting membrane potential V_m (=E_m) of the non-clamped fibroblasts. When freshly isolated atrial fibroblasts were stretched, the I-V curve was bended upwardly, and intersected the voltage axis at a more negative V_m (Fig. 3A). In contrast, mechanical compression of isolated fibroblasts with a glass stylus caused downward flection of the I-V curve and shifted V_m to more positive values (Fig. 3C). Both, the compression-stimulated and the stretch-reduced currents reversed their polarity close to 0 mV, as would be expected for a mechanosensitive, non-selective cation channel that can conduct Na⁺, K⁺ and Cs⁺ ions [21, 23]. The compression-induced current was inhibited by low concentrations of gadolinium (8 μM; Fig. 4), a compound, which is commonly used to block current flow through mechanosensitive ion channels [21, 23].

Figure 3

Mechanosensitivity of membrane currents in fibroblasts that had been freshly isolated from rat atria. Current-voltage relations (I-V curves, current is measured at the end of the 140 ms clamp pulses). A: I-V curves before (empty triangles) and during (more...)

Figure 4

Inhibition of compression-induced currents in freshly isolated fibroblasts from rat atria with gadolinium (8 μM). Current-voltage relations (I-V curves, the current was measured at the end of 140 ms clamp pulses). I-V curve before (empty triangles) (more...)

Similar to the effect of mechanical stretch, gadolinium hyperpolarized the membrane potential of the fibroblasts, albeit to a larger extent (Figs. 2 and 4). The voltage-dependence and the gadolinium-sensitivity of the compression-induced currents resembled the stretch-activated currents in cardiomyocytes [18, 20, 22, 25, 67, 68] and other cell types (for review of stretch-activated, non-selective cation channels (SACs) see reference [61]). Hence, we conclude that modulation of MSC is responsible at least in part for the mechanosensitivity of cardiac fibroblasts.

The findings obtained with freshly isolated single fibroblasts were similar to those reported for multicellular preparations. However, fibroblasts that were embedded in their normal three-dimensional tissue environment differed in some aspects from isolated cells. For example, fibroblasts in multi-cellular preparations had a lower E_m of approximately -22 mV with a wide distribution range between -5 and -70 mV (Fig. 5).

Figure 5

Distribution of the membrane potential of mechanosensitive fibroblasts in the rat sinus node region. (From Kiseleva et al. [34], with permission from Elsevier).

The wide distribution range of E_m suggests that fibroblast in their normal environment are exposed to variable mechanical stretch or compression, even in the pause between the contractions and relaxations of the myocardium. A similar distribution of E_m of atrial fibroblasts was found in non-contracting tissue. A step-by-step increase of the amount of stretch gradually hyperpolarized E_m only in 80% of the fibroblasts (Fig. 6) and reduced the frequency of spontaneous contractions of the myocardial tissue preparations [26, 28].

Figure 6

Specific reaction of the membrane potential of a rat atrial fibroblast to mechanical stretch of the tissue. Synchronous registration of the mechanogram (top curve) and mechanically induced potentials (MIP) of the fibroblast (bottom curve). The symbol (↑) (more...)

At the same time, a step-by-step increase of stretch of the tissue depolarized E_m in 20% of the fibroblasts (Fig. 7). This latter reaction to artificial stretch results from compression of fibroblasts, which could explain why in situ during tissue contraction (systole) not all of the fibroblasts are depolarized, and during tissue relaxation (diastole) not all of fibroblast are hyperpolarized. In our review we will discuss the typical response of fibroblasts – depolarization during compression and hyperpolarization during stretch.

Figure 7

A stepwise increase of the stretch intensity in 4-s intervals gradually raised the resting force (RF) of the preparations. This effect was paralleled by membrane depolarization and reduced MIP amplitudes in approximately 20% of the atrial fibroblasts (more...)

Most fibroblasts are probably compressed during the contractions, which will depolarize their membrane potential, thereby eliciting so-called "mechanically induced potentials" (MIPs) [32]. MIPs in cardiac fibroblasts differ from action potentials in cardiomyocytes by their lack of a fast upstroke velocity and overshoot. They usually start with a delay of several milliseconds after the onset of action potentials [26, 34, 36, 39]. MIPs can be detected only in contracting atrial tissue and strictly follow the rhythm of contractions. Upon hyperpolarization of the resting potential, the amplitude of the MIPs increases, whereas membrane depolarization will decrease the MIP amplitudes.

It is a major challenge to analyze the pathways, through which mechaniccal compression may activate ion channels in the plasma membrane of atrial fibroblasts. One possibility would be that the mechanical energy is transmitted through changed tension of the lipid bilayer [61]. Alternatively, the cytoskeleton may conduct mechanical stress from the site of physical deformation to the channel protein [30, 61]. The latter possibility is supported by recent observations with multicellular preparations, where disruption of cytoskeletal proteins - through depolymerisation of F-actin with cytochalasin D and tubulin degradation with colchicine - decreased the MIP amplitudes [30]. This effect of cytochalasin D on the MIP amplitudes could be attenuated by intracellular application of Mg-ATP to promote actin polymerisation [5]. Thus, similar to the findings made with other stretch-activated ion channels and different types of voltage-gated channels [16, 19, 45], the cytoskeleton appears to play a major role in the signal transfer from the site of mechanical deformation to the ion channels in the plasma membrane of atrial fibroblasts. Cytochalasin D and colchicine did not fully suppress the MIPs in atrial fibroblasts suggesting that cytoskeletal disruption by these drugs was either incomplete, or that additional, cytoskeleton-independent mechanisms are involved in the regulation of mechanically operated ion channels in atrial fibroblasts [61].

Age dependence of the electrical properties of cardiac fibroblasts

Cardiac arrhythmia is a clinical disposition that affects mainly the older population. Since fibroblasts in the heart have been implicated in the pathophysiology of cardiac arrhythmia, it is important to explore whether the mechano-electrical properties of the cells would change with increasing age of the individuals. In a first attempt to resolve this issue, we analyzed the electrophysiological characteristics of fibroblasts from the right atria of young (3 months) and old (15 months) healthy rats. In both groups, the membrane potential of individual fibroblasts in multicellular right atrial preparations varied between -5 mV and -70 mV. However, several interesting differences in the electrophysiological properties between cardiac fibroblasts from 3 and 15 months old animals were detected.

For example, the distribution of resting membrane potential values varied remarkably between young (Fig. 8A) and old (Fig. 8B) rats. The majority of atrial fibroblasts in young rats had resting membrane potentials between -10 mV and -15 mV (n=124). For comparison, E_m of cardiac fibroblasts from old rats was mostly in the range from -20 mV to -25 mV (n=162). The mean values of the fibroblast membrane potentials were 19.23 ± 1.02 mV and 29.41 ± 1.23 mV in young and old animals, respectively.

Figure 8

Distribution of the resting membrane potentials of right atrial fibroblasts from 3 months old (A) and 15 months old (B) rats.

The membrane resistance of atrial fibroblasts from young rats was usually between 0.4 GΩ and 0.6 GΩ, whilst older animals had membrane resistances between 0.7 and 0.9 GΩ. Thus, the higher membrane potential of fibroblasts from old rats was associated with increased membrane resistances.

The following experiments were performed on fibroblasts from human right atrial trabeculae. The tissue biopsies were obtained from patients between 65 and 78 years of age, who underwent aorto-coronary bypass surgery. The right atrial function of the patients was clinically normal. The resting membrane potential of these fibroblasts was -15.9 ± 2.1 mV and the membrane resistance 4.1 ± 0.1 GΩ. [26].

Altered electrical function of cardiac fibroblasts after myocardial infarction

Contractile failure of the heart due to cardiac arrhythmia is a major cause for cardiovascular mortality. The risk for arrhythmia is particularly high after myocardial infarction [6, 37, 51, 54, 55, 66]. The mechanisms that underlie the electrical instability of the ischemic heart are not well understood. Recent findings suggest that a phenomenon, which is commonly referred to as mechano-electric feedback, might play an important role in cardiac arrhythmia [8, 15, 52, 56, 65]. Mechano-electric feedback refers to the situation that mechanical changes in the myocardium, which may result from variable filling pressures, can modulate the electrical function of the heart. This, in turn, may affect the spontaneous electrical activity of the myocardium [44].

The use of microelectrode recordings and the patch-clamp technology made it possible to analyze in detail membrane potential changes of cardiac cells in response to mechanical stretch. Electrophysiological studies revealed that mechano-electric coupling, i.e. the electrical response of cardiomyocytes to stretch, involves the transmembrane influx of cations through stretch-activated channels [18, 20, 22, 25, 67, 68]. Correspondingly, the influx of cations into cardiomyocytes through stretch-activated channels was enhanced and extra action potentials occurred under ischemia, which could possibly elicit cardiac arrhythmia [9, 10, 12, 20, 25, 49, 50, 52]. The sensitivity of the cardiac myocytes to mechanical stretch was enhanced after infarction, suggesting increased expression levels of stretch-activated ion channels during myocardial remodeling after infarction [20, 27, 35].

In addition to the electrophysiological changes of cardiomyocytes, fibroblasts in the heart showed pronounced hyperpolarization of their resting membrane potential (Fig. 9) and increased sensitivity to mechanical stretch after infarction. Modulation of the electrical function of the atrial fibroblasts may destabilize the rhythmical activity of the ischemic myocardium. This idea is supported by the observation that mechanical stretch hyperpolarized E_m of atrial fibroblasts to a higher extent in tissue specimens from infarcted (Fig. 10) than from healthy rat hearts [29]. Furthermore, the susceptibility of the membrane potential of the fibroblasts to mechanical stimulation correlated positively with the infarct size, and artificial stretching of the cells reduced the frequency of spontaneous contractions [29]. While these findings document a role of atrial fibroblasts in the mechano-electric function of normal and diseased hearts, it remains to be clarified as to how electrical signals that are generated by the atrial fibroblasts can modulate the spontaneous contractile activity of the myocardium. An important clue to this issue may come from a detailed analysis of the pathways, through which fibroblasts and myocytes in the heart communicate with each other. In the following paragraphs we will discuss the current concepts about the transfer of electrical signals between fibroblasts and myocytes in the heart.

Figure 9

Frequency distribution of the resting membrane potential of fibroblasts from rat atria 20 days after experimental myocardial infarction. (A) shows the data from rats with small infarct sizes of 16.5 ± 0.6 % of the left endocardial circumference (more...)

Figure 10

Response of the membrane potential of a right atrial fibroblast to long lasting artificial stretch of the tissue 20 days after experimental myocardial infarcttion (approx. 40 % of the left endocardial circumference). Note the excessive hyperpolarization (more...)

Electrical coupling between fibroblasts and myocytes

Electrical communication among cardiac fibroblasts

Electrical communication of fibroblasts in the heart was studied in right atrial tissue preparations from various species with the use of microelectrode recording techniques. A bidirectional transcellular electrical current flow could be monitored in 65% of the fibroblasts from frog hearts when two microelectrodes were randomly inserted into different cells at a distance of ≈40 μm [24, 36, 39]. Artificial hyperpolarization of one fibroblast shifted the membrane potential of the second cell to more negative values and vice versa (Fig. 11A, 11C). By consequence, the MIP amplitudes were increased not only in the polarized cell, but also in the fibroblast, to which the hyperpolarizing current pulses were conducted (Fig. 11A, 11C). Conversely, depolarization of one fibrobroblast shifted the membrane potential of the second cell to less negative values, and this effect was also bidirectional (Fig. 11B, 11D). The transmission coefficient was 0.11 ± 0.02 mV and did not depend on the magnitude of artificial intracellular polarization [36].

Figure 11

Bidirectional electrical coupling of two fibroblasts in the right atrium of a frog. The two microelectrodes were located 40 μm apart from each other. A – artificial intracellular hyperpolarization of one cell (1) shifted the membrane potential (more...)

Intracellular injection of Lucifer Yellow into a cardiac fibroblast resulted in spreading of the dye to other fibroblasts through gap junctions (Fig. 12).

Figure 12

Light photomicrograph demonstrating the intercellular transfer of Lucifer Yellow between to fibroblasts through gap junctions. Scale bar indicates 100 μm. (From Kiseleva et al. [32], with permission from European Society of Cardiology).

Furthermore, recent data demonstrated the existence of two populations of fibroblasts in the sinoatrial node region of rabbit hearts; one, which expresses the gap junction protein connexin (Cx) 45, and a second type of fibroblast, which is positive for Cx40. Additionally, some fibroblasts in the heart were reported to contain Cx43 [4].

In approximately 20% of the cells, monodirectional coupling between two fibroblasts at a distance of ≈40μm was observed [33]. In this case, artificial polarization of one cell caused a shift of the membrane potential of the second cell, but not vice versa. Monodirectional current flow between two cardiac fibroblasts can be explained on the basis of variable input resistances of the cells. For example, in two adjacent cells with different input resistances, any change of the membrane potential of the cell with low input resistance would cause a similar membrane potential shift of the cell with high input resistance. However, due to its lower input resistance, the first cell would be less susceptible to membrane potential changes, which are transmitted from the other fibroblasts. Notably, the input resistance of a cell is dependent not only on its size, but also on its three-dimensional shape and, in particular, on its electrical coupling to neighboring cells. Monodirectional current flow between two cardiac fibroblasts can therefore result from different input resistances due to asymmetrical coupling of their cell membrane with adjacent cardiomyocytes.

Electrical interaction between fibroblasts and myocytes in the heart

A detailed knowledge of the intercellular routes, through which fibroblasts and myocytes in the heart communicate to each other is necessary. Several attempts to demonstrate electrotonic interaction of atrial fibroblasts and cardiomyocytes via gap junctions were not successful. Electron microscopic studies showed clearly, that the plasma membranes of fibroblasts and myocytes in the sinoatrial node region form close contacts [29, 34, 48], but without typical gap junction structures visible [48]. We therefore assumed that single gap junction channels rather than clusters of cell-cell contacts might exist in the cell membrane of atrial fibroblasts and cardiac myocytes [40]. Unfortunately, single gap junction channels were difficult to detect by means of standard morphological methods [29, 34, 48]. More recently, the gap junction protein Cx45 was identified in the sinus node region of rabbit hearts at sites where the plasma membranes of cardiomyocytes and atrial fibroblasts formed close contacts [4]. Furthermore, intracellular injection of Lucifer Yellow into cardiac fibroblasts resulted in lateral spreading of the dye along extended threads of interconnected fibroblasts and, in particular, between neighboring fibroblasts and cardiomyocytes in the sinoatrial node. These findings strongly suggest that electrical signals, which may arise in cardiac fibroblasts upon mechanical stimulation, can be transmitted to the adjacent cardiomyocytes [4].

Electrical interaction of myocytes and fibroblasts in the heart via gap junctions is also supported by the results of intracellular microelectrode studies. As discussed earlier, gap junctions constitute an important pathway for the transfer of action potentials from cardiac myocytes to adjacent fibroblasts [57, 58, 59]. Isolated fibroblasts had a E_m of approximately -37 mV, whereas the resting membrane potential of isolated cardiomyocytes was normally in the range of -80 mV. Cultures that were enriched in cardiac fibroblasts had membrane potentials (E_m) between -10 and -20 mV, while the membrane potential of isolated myoblasts was approximately -60 mV. Electrical coupling of the two cell types in mixed cultures is indicated by a depolarization of the membrane potential of the myoblasts to -50 mV and a shift of E_m of the co-cultured fibroblasts to more negative values between -40 and -60 mV.

Due to extensive gap junctional coupling, fibroblasts in the myocardial tissue had a E_m, which is close to the resting membrane potential of the cardiac myocytes (approx. -80 mV). Intracellular microelectrode recordings demonstrated that membrane potential changes in fibroblasts in response to the action potentials in cardiac myocytes occurred with a much slower upstroke velocity (delayed time course) in the myocardial tissue (Fig. 13).

Figure 13

Effect of intracellular de- and hyperpolarization on the action potential in a cardiomyocyte (A1, B1, C1 and D1) and the membrane potential of a rat cardiac fibroblast (A2, B2, C2 and D2). See text for more details (Modified from Kohl et al. [40] with (more...)

In the experiment, which is shown in Fig. 13, both cell types could be distinguished from each other by their different responses to de- and hyperpolarizing intracellular current injections (10^-9 A). Intracellular polarization of this small current does not lead to changes in action potential (AP) configuration of the cardiomyocytes, because all cardiomyocytes have input resistances, which are below the microelectrode resistance. In contrast, the input resistance of fibroblasts is much higher than the microelectrode resistance, and therefore polarization of a fibroblast will lead to changes in MIP amplitude. A typical AP registered from a cardiomyocyte is shown in Fig. 13A1, whereas Fig. 13A2 depicts the AP-like potential recorded in a fibroblast. Intracellular hyperpolarization of the cardiac myocyte did not change the shape of its AP due to the low input resistance of the cell (Fig. 13B1). Hyperpolarization of the fibroblast membrane also had no impact on the amplitude of the AP-like potential, but resulted in a delayed repolarization of the membrane potential instead (Fig. 13B2). This "plateau phase" corresponds to the mechanically induced potential (MIP), which is unmasked from the AP-like potential upon hyperpolarization of the membrane potential. Depolarizing current injections caused a transient hyperpolarization of the membrane potential of the cardiac fibroblast right after the AP transmitted from the cardiac myocyte (Fig. 13C2). Depolarization of the cardiac myocyte did not affect the time course of the AP in this cell (Fig. 13C1). Moreover, the upstroke velocity of the AP (Fig. 13D1) is typical for cardiomyocytes. The AP-like potential (Fig. 13D2) never reaches the upstroke velocity of the AP depolarization since it is much slower than that.

Intracellular injection of Lucifer Yellow into a cell with AP-like potentials (Fig. 13 A2, B2, C2, D2) demonstrates that it is cardiac fibroblast (A. Kamkin and I. Kiseleva, 1996. Unpublished data).

These results provide further evidence that myocytes and fibroblasts in the heart are electrically coupled through gap junctions [40].

Role of cardiomyocyte–fibroblast coupling in healthy and diseased hearts

Cardiac fibroblasts and myocytes are cellular components of the mechano-electric feedback mechanism in the heart. Fibroblasts can change their membrane potential in response to mechanical deformation of the plasma membrane such that contraction of the myocardium compresses the fibroblasts and depolarizes their membrane potential. These contraction-evoked membrane depolarizations are referred to as mechanically induced potentials (MIPs) and result from a transmembrane influx of cations, mainly of sodium ions, through mechanosensitive channels in the plasma membrane of these cells [21, 23]. On the contrary, hyperpolarization of the fibroblasts occurs in response to channel inactivation during mechanical distension of the cells [21, 23]. Hence, the mechanosensitive channels in cardiac fibroblasts operate as stretch-inactivated channels (SICs).

In addition to SIC, cardiac fibroblasts may also contain voltage-dependent potassium K_v channels in their plasma membrane [21, 23]. Voltage-sensitive potassium currents were monitored in freshly isolated cardiac fibroblasts (Fig. 1A and 2A). Furthermore, the compression-activated conductance followed an outwardly rectifying voltage dependence.

Unlike the cardiac fibroblasts, cardiomyocytes become depolarized by mechanical stretch. Stretch-induced depolarization of their membrane potential is determined by the influx of cations through SACs [18, 20, 22, 25, 67, 68]. When the membrane depolarization reaches threshold, extra action potentials can occur, which potentially give rise to arrhythmic myocardial contractions [27]. Thus, mechanical stretch exerts opposite effects on the membrane potential of the cardiomyocytes and fibroblasts: it depolarizes the membrane potential of the cardiomyocytes and hyperpolarizes the fibroblast membrane.

Electrical communication between cardiac myocytes and fibroblasts is based on the presence of Cx45-positive gap junction channels. Camelliti et al. estimated that approximately 10% of the overall Cx45 protein content in the heart is localized to the sites of contact formation between fibroblasts and myocytes [4]. Since the input membrane resistance is significantly higher in fibroblasts than in the cardiomyocytes, even a relatively small number of Cx45 molecules would be sufficient to maintain the electrical interaction between myocytes and fibroblasts in the heart.

Mathematical modeling was performed for a cell pair consisting of a single myocyte from the sino-atrial node weakly coupled by 30 gap junction channels to a mechanosensitive fibroblast. Theoretical analysis predicted that depolarization of E_m of the fibroblast would decrease the duration of slowly depolarizing voltage trajectory of the pacemaker potential (between action potentials) of adjacent sino-atrial node myocytes, thereby increasing the frequency of spontaneous atrial contractions [40, 41]. On the contrary, hyperpolarization of E_m of the fibroblast would increase the time of slowly depolarizing voltage trajectory of the pacemaker potential of adjacent sino-atrial node myocyte. This, in turn, would decrease the frequency of spontaneous atrial contractions.

It is therefore likely that MIPs have a role in the mechano-electric feedback in the heart. The depolarization phase of the MIP does not influence the AP, because it occurs during the absolute refractory period of the cardiomyocytes. However, repolarization of the MIP can possibly affect the duration of the repolarization phase of the AP at the level of APD50 or APD90 (action potential duration at 50% and 90% of repolarization). Repolarization of the MIP together with stretch-induced depolarization of AP can increase APD50 or APD90. As a consequence, increases of APD50 and APD90 may provide a risk factor for the occurrence of arrhythmias.

Tissue remodeling after myocardial infarction is associated with profound changes in both, the number and the electrical properties of cardiac fibroblasts. For example, the resting membrane potential of rat atrial fibroblasts increased in parallel to the infarct size after coronary artery occlusion. Furthermore, atrial fibroblasts from infarcted rat hearts were approximately 10-times more sensitive to mechanical deformation of their plasma membrane and responded with a stronger hyperpolarization of their membrane potential than fibroblasts from healthy hearts [29].

Furthermore, remodeling after myocardial infarction also increased the sensitivity of the cardiomyocytes to mechanical stretch. Thus, very small amounts of stretch were sufficient to evoke membrane depolarization of the cardiomyocytes in the post-ischemic heart. Mechanically induced membrane depolarization resulted from activation of mechanosensitive ion channels in the plasma membrane of the cardiomyocytes. These channels exhibit a selective permeability for cations, mainly for sodium ions [18, 20, 22, 25, 67, 68]. Extra action potentials appeared, when the mechanically induced depolarizations reached a threshold. As a consequence, extra-systoles and, in the worst case, fibrillations can occur [18, 20, 22, 25, 68].

Focal disorganization of gap junctions and down-regulation of Cx43, a major cardiac gap junction protein, are typical features of myocardial remodeling. These processes may also play a permissive role for the development of arrhythmia in human cardiomyopathies [42, 43]. Abnormal localization of Cx43 was often observed on lateral surfaces of surviving cardiomyocytes after myocardial infarction in rats [46]. Cx43 was redistributed from intercalated discs to the lateral surface of structurally compromised myocytes [3]. These findings suggest that ischemia caused ectopic expression of Cx43, which could be responsible for abnormal electrical conductivity around the infarcted area [47]. A changing pattern of connexin expression in sheep was caused by the invasion of Cx45-positive fibroblasts into the ischemic zone within 24 h after infarction. The number of Cx45-expressing fibroblasts reached a maximum after 6 days. It is important to note that the sensitivity of the cardiac fibroblasts to mechanical stretch was highest 7 days after myocardial infarction due to coronary artery occlusion [29]. From thereon, the expression of Cx43 increased, whereas Cx45 expression was reduced [3].

The rapid infiltration of the damaged tissue with highly coupled fibroblasts likely interferes with the excitation control and the spreading of electrical signals from the infarction border zone to the surrounding tissue. In addition to the enhanced mechanosensitivity of cardiac fibroblasts and myocytes after infarction, altered electrical interaction between the two cell types may increase the risk for contractile dysfunction. This possibility is in agreement with the finding that the nonspecific gap junction channel uncoupler heptanol limited necrosis and infarct size [17, 63].

Conclusions and perspectives

A detailed knowledge of the trafficking routes, through which cardiac myocytes and fibroblasts communicate with each other is necessary for the understanding of normal and impaired cardiac function. Interaction of fibroblasts and myocytes involves the intercellular exchange of electrical signals via gap junctions [3, 4]. Gap junction channels were demonstrated with the use of electrophysiological and morphological techniques at the sites of contact formation between cardiomyocytes and fibroblasts. Gap junctions between atrial fibroblasts and cells of the sinoatrial node may provide the structural correlate for the transmission of membrane potential changes from the fibroblasts to the cardiac pacemaker cells. Through this pathway, an increase in wall distension of the right atrium, i.e. due to enhanced venous return, may elicit a chronotropic response of the heart [40]. This may have serious implications for the contractile activity of the heart resulting in a decrease of heart rates, when hyperpolarization of the fibroblasts exceeds depolarization of the cardiomyocytes. Furthermore, extra systoles and cardiac arrhythmia may occur, when mechanically induced membrane depolarization of the cardiac myocytes predominates.

The gap junctional conductivity can be regulated by protein phosphorylation involving different protein kinases, i.e. protein kinase A, C, and G as well as tyrosine kinases and mitogen-activated protein kinase. Some studies have demonstrated that activation of cAMP-dependent protein kinase A can enhance gap junction coupling in Cx43-positive cardiomyocytes. The conductance of gap junctions can also be regulated by small ions like H⁺, Ca²⁺, Mg²⁺ and Na⁺. An increase in the intracellular concentration of Na⁺ may reduce the gap junction conductance. A drop of the intracellular pH can also reduce gap junctional conductivity, with the histidine residue 95 acting as a pH sensor in Cx43 channels (for review see reference [11]).

In addition to gap junction regulation by extracellular mechanical forces, there is a close relation between gap junctions and adhesion junctions and their linkage to the cytoskeleton. This can be inferred from experiments on neoformation of cell-to-cell coupling, concomitant upregulation of adherens and gap junctions after mechanical stretch, and human cardiomyopathies caused by genetic defects in cell-cell adhesion junction proteins. The molecular mechanisms responsible for the interaction between mechanical and functional cell-to-cell coupling remain to be elucidated (for review see reference [62]).

It is likely that stretch-activated changes in cardiac myocyte structure and function are mediated by signaling pathways that are initiated by interactions between integrins and extracellular matrix proteins. Indeed, overexpression of β₁ integrin was sufficient to induce a hypertrophic response in cultured neonatal rat ventricular myocytes and enhanced the effects of β₁ adrenergic stimulation [60].

Future research will need to elucidate the complete signaling pathway by which mechanical stimulation of cardiac myocytes alters the gap junctional communication of cardiac cells. A detailed analysis of the molecular signaling mechanisms between cardiac myocytes and fibroblasts may eventually allow one to develop new strategies for the treatment of cardiac arrhythmias.

Acknowledgements

This study was supported by the Alexander von Humboldt-Stiftung, the Deutsche Forschungsgemeinschaft (DFG), and a travel grant from the Humboldt-University (Charité).

References

1.

Adler CP, Ringlage WP, Bohm N. DNS-Gehalt und Zellzahl in Herz und Leber von Kindern. Pathol Res Pract. (1981);172:25–41. (German). [PubMed: 7312710]

2.

Anversa P, Olivetti G, Melissari M, Loud AV. Stereological measurement of cellular and subcellular hypertrophy and hyperplasia in the papillary muscle of adult rat. J Mol Cell Cardiol. (1980);12:781–795. [PubMed: 6448301]

3.

Camelliti P, Devlin GP, Matthews KG, Kohl P. Spatially and temporally distinct expression of fibroblast connexins after sheep ventricular infarction. Cardiovasc Res. (2004);62(2):415–425. [PubMed: 15094361]

4.

Camelliti P, Green CR, LeGrice I, Kohl P. Fibroblast network in rabbit sinoatrial node: structural and functional identification of homogeneous and heterogeneous cell coupling. Circ Res. (2004);94(6):828–835. [PubMed: 14976125]

5.

Chhabra D, Nosworthy NJ, dos Remedios CG. The role of ATP, ADP and divalent cations in the formation of binary and ternary complexes of actin, cofilin and DNase I. Electrophoresis. (2000);21:3863–3869. [PubMed: 11271505]

6.

D'Alonzo AJ, Sewter JC, Darbenzio RB, Hess TA. Effects of dofetilide on electrical dispersion and arrhythmias in post-infarcted anesthetized dogs. Basic Res Cardiol. (1995);90(5):424–434. [PubMed: 8585865]

7.

Davies MJ, Pomerance A. Quantitative study of ageing changes in the human sinoatrial node and internodal tracts. British Heart Journal. (1972);34:150–160. [PMC free article: PMC487110] [PubMed: 5007792]

8.

Dean JW, Lab MJ. Arrhythmia in heart failure: role of mechanically induced changes in electrophysiology. Lancet. (1989);1(8650):1309–1312. (Review). [PubMed: 2566835]

9.

Dean JW, Lab MJ. Effect of changes in load on monophasic action potential and segment length of pig heart in situ. Cardiovasc Res. (1989);23:887–896. [PubMed: 2620316]

10.

Dean JW, Lab MJ. Regional changes in myocardial refractoriness during load manipulation in the in-situ pig heart. J Physiol. (1990);429:387–400. [PMC free article: PMC1181706] [PubMed: 2277353]

11.

Dhein S. Gap junction channels in the cardiovascular system: pharmacological and physiological modulation. Trends Pharmacol Sci. (1998);19(6):229–241. [PubMed: 9666714]

12.

Dilly SG, Lab MJ. Electrophysiological alternans and restitution during acute regional ischaemia in myocardium of anaesthetized pig. J Physiol (London). (1988);402:315–333. [PMC free article: PMC1191893] [PubMed: 3236241]

13.

Eghbali M, Czaja MJ, Zeydel M, Weiner FR, Zern MA, Seifter S, Blumenfeld OO. Collagen chain mRNAs in isolated heart cells from young and adult rats. J Mol Cell Cardiol. (1988);20:267–276. [PubMed: 3398057]

14.

Feld Y, Melamed-Frank M, Kehat I, Tal D, Marom S, Gepstein L. Electrophysiological modulation of cardiomyocytic tissue by transfected fibroblasts expressing potassium channels: a novel strategy to manipulate excitability. Circulation. (2002);105(4):522–529. [PubMed: 11815438]

15.

Franz MR. Mechanoelectrical feedback in ventricular myocardium. Cardiovasc Res. (1996);32(1):15–24. (Review) [PubMed: 8776399]

16.

Galli A, DeFelice LJ. Inactivation of l-type Ca channels in embryonic chick ventricular cells: dependence on the cytoskeletal agents colchicine and taxol. Biophys J. (1994);67:2296–2304. [PMC free article: PMC1225614] [PubMed: 7696470]

17.

Garcia-Dorado D, Rodriguez-Sinovas A, Ruiz-Meana M. Gap junction-mediated spread of cell injury and death during myocardial ischemia-reperfusion. Cardiovasc Res. (2004);61:386–401. [PubMed: 14962471]

18.

Isenberg G, Kazanski V, Kondratev D, Gallitelli MF, Kiseleva I, Kamkin A. Differential effects of stretch and compression on membrane currents and [Na⁺]_c in ventricular myocytes. Prog Biophys Mol Biol. (2003);82(13):43–56. (Review) [PubMed: 12732267]

19.

Johnson BD, Byerly L. A cytoskeletal mechanism for Ca²⁺ channel metabolic dependence and inactivation by intracellular Ca²⁺ Neuron. (1993);10:797–804. [PubMed: 8098608]

20.

Kamkin A, Kiseleva I, Isenberg G. Stretch-activated currents in ventricular myocytes: amplitude and arrhythmogenic effects increase with hypertrophy. Cardiovasc Res. (2000);48:409–420. [PubMed: 11090836]

21.

Kamkin A, Kiseleva I, Isenberg G. Activation and inactivation of a non-selective cation conductance by local mechanical deformation of acutely isolated cardiac fibroblasts. Cardiovasc Res. (2003);57(3):793–803. [PubMed: 12618241]

22.

Kamkin A, Kiseleva I, Isenberg G. Ion selectivity of stretch-activated cation currents in mouse ventricular myocytes. Pflügers Arch. (2003);446(2):220–231. [PubMed: 12739160]

23.

Kamkin A, Kiseleva I, Isenberg G, Wagner KD, Günther J, Theres H, Scholz H. Cardiac fibroblasts and the mechanoelectric feedback mechanism in healthy and diseased hearts. Prog Biophys Mol Biol. (2003);82(13):111–120. [PubMed: 12732272]

24.

Kamkin A, Kiseleva I, Kircheis R, Kositzky G. Bioelectric activity of frog atrium cells with non-typical impulse activity. Abhandlungen der Akademie der Wissenschaften der DDR (Abteilung Mathematik - Naturwissenschaft - Technik). (1988);1:103–106.

25.

Kamkin A, Kiseleva I, Wagner KD, Bohm J, Theres H, Günther J, Scholz H. Characterization of stretch-activated ion currents in isolated atrial myocytes from human hearts. Pflügers Arch. (2003);446(3):339–346. [PubMed: 12799902]

26.

Kamkin A, Kiseleva I, Wagner KD, Lammerich A, Bohm J, Persson PB, Günther J. Mechanically induced potentials in fibroblasts from human right atrium. Exp Physiol. (1999);84:347–356. [PubMed: 10226175]

27.

Kamkin A, Kiseleva I, Wagner KD, Leiterer KP, Theres H, Scholz H, Günther J, Lab MJ. Mechanoelectric feedback in right atrium after left ventricular infarction in rats. J Mol Cell Cardiol. (2000);32:465–477. [PubMed: 10731445]

28.

Kamkin A, Kiseleva I, Wagner KD, Lozinsky I, Günther J, Scholz H. Mechanically induced potentials in atrial fibroblasts from rat hearts are sensitive to hypoxia/reoxygenation. Pflügers Arch. 2003;446(2):169–174. [PubMed: 12739154]

29.

Kamkin A, Kiseleva I, Wagner KD, Pylaev A, Leiterer KP, Theres H, Scholz H, Günther J, Isenberg G. A possible role for atrial fibroblasts in postinfarction bradycardia. Am J Physiol. (2002);282:H842–H849. [PubMed: 11834477]

30.

Kamkin A, Kiseleva I, Wagner KD, Scholz H, Theres H, Kazanski V, Lozinsky I, Günther J, Isenberg G. Mechanically induced potentials in rat atrial fibroblasts depend on actin and tubulin polymerisation. Pflügers Arch. (2001);442:487–497. [PubMed: 11510879]

31.

Kamkin A, Kohl P, Kiseleva I, Pylaev A. Capacitive interaction between cardiomyocytes and fibroblasts across their basal membrane. Pflügers Arch. (1992);420(suppl. II):R84–332. (Abstract).

32.

Kiseleva I, Kamkin A, Kohl P, Lab M. Calcium and mechanically induced potentials in fibroblasts of rat atrium. Cardiovasc Res. (1996);32:98–111. [PubMed: 8776407]

33.

Kiseleva I, Kamkin A, Leiterer KP, Kohl P. Interaction of mechanosensitive cells with surrounding cells in the right atrium of the rat heart. J Mol Cell Cardiol. (1993);25(suppl.1):S76. (Abstract)

34.

Kiseleva I, Kamkin A, Pylaev A, Kondratjev D, Leiterer KP, Theres H, Wagner KD, Persson PB, Günther J. Electrophysiological properties of mechanosensitive atrial fibroblasts from chronic infarcted rat heart. J Mol Cell Cardiol. (1998);30(6):1083–1093. [PubMed: 9689583]

35.

Kiseleva I, Kamkin A, Wagner KD, Theres H, Ladhoff A, Scholz H, Günther J, Lab MJ. Mechanoelectric feedback after left ventricular infarction in rats. Cardiovasc Res. (2000);45:370–378. [PubMed: 10728357]

36.

Kiseleva IS, Kamkin AG, Kircheis R, Kositski GI. Intercellular electrotonical interaction in the cardiac sinus node in the frog. Reports of Academy of Science of USSR. (1978);292(6):1502–1505. (Russian). [PubMed: 3829907]

37.

Kleber AG, Fast V. Molecular and cellular aspects of re-entrant arrhythmias. Basic Res Cardiol. (1997);92(Suppl 1):111–119. (Review). [PubMed: 9202851]

38.

Kohl P. Heterogeneous cell coupling in the heart an electrophysiological role for fibroblasts. Circ Res. (2003);93:381–383. [PubMed: 12958139]

39.

Kohl P, Kamkin A, Kiseleva I, Streubel T. Mechanosensitive cells in the atrium of frog heart. Exp Physiol. (1992);77:213–216. [PubMed: 1543586]

40.

Kohl P, Kamkin AG, Kiseleva IS, Noble D. Mechanosensitive fibroblasts in the sino-atrial node region of rat heart: interaction with cardiomyocytes and possible role. Exp Physiol. (1994);79:943–956. [PubMed: 7873162]

41.

Kohl P, Noble D. Mechanosensitive connective tissue: potential influence on heart rhythm. Cardiovasc Res. (1996);32:62–68. [PubMed: 8776404]

42.

Kostin S, Dammer S, Hein S, Klovekorn WP, Bauer EP, Schaper J. Connexin 43 expression and distribution in compensated and decompensated cardiac hypertrophy in patients with aortic stenosis. Cardiovasc Res. (2004);62(2):426–436. [PubMed: 15094362]

43.

Kostin S, Rieger M, Dammer S, Hein S, Richter M, Klovekorn WP, Bauer EP, Schaper J. Gap junction remodeling and altered connexin43 expression in the failing human heart. Mol Cell Biochem. (2003);242(12):135–144. [PubMed: 12619876]

44.

Lab MJ. Is there mechanoelectric transduction in cardiac muscle? The monophasic action potential of the frog ventricle during isometric and isotonic contraction with calcium deficient perfusions. S Afr J Med Sci. (1968);33:60. (Abstract).

45.

Maltsev VA, Undrovinas AI. Cytokleleton modulates coupling between availability and activation of cardiac sodium channel. Am J Physiol. (1997);273:H1832–H1840. [PubMed: 9362250]

46.

Matsushita T, Oyamada M, Fujimoto K, Yasuda Y, Masuda S, Wada Y, Oka T, Takamatsu T. Remodeling of cell-cell and cell-extracellular matrix interactions at the border zone of rat myocardial infarcts. Circ Res. (1999);85(11):1046–1055. [PubMed: 10571536]

47.

Matsushita T, Takamatsu T. Ischaemia-induced temporal expression of connexin43 in rat heart. Virchows Arch. (1997);431(6):453–458. [PubMed: 9428934]

48.

Maziere de AMGL, Ginneken van ACG, Wilders R, Jongsma HJ, Bouman LN. Spatial and functional relationship between myocytes and fibroblasts in the rabbit sinoatrial node. J Mol Cell Cardiol. (1992);24:567–578. [PubMed: 1518074]

49.

Murphy CF, Horner SM, Dick DJ, Coen B, Lab MJ. Electrical altemans and the onset of rate-induced pulsus alternans during acute regional ischaemia in the anaesthetised pig heart. Cardiovasc Res. (1996);32:138–147. [PubMed: 8776411]

50.

Murphy CF, Lab MJ, Horner SM, Dick DJ, Harrison FG. Regional electromechanical alternans in anesthetized pig hearts: modulation by mechanoelectric feedback. Am J Physiol. (1994);267:H1726–1735. [PubMed: 7977805]

51.

Naccarella F, Lepera G, Rolli A. Arrhythmic risk stratification of post-myocardial infarction patients. Curr Opin Cardiol. (2000);15:1–6. (Review). [PubMed: 10666655]

52.

Nazir SA, Lab MJ. Mechanoelectric feedback and atrial arrhythmias. Cardiovasc Res. (1996);32(1):52–61. (Review). [PubMed: 8776403]

53.

Opthof T, deJonge B, Jongsma HJ, Bouman LN. Functional morphology of the mammalian sinuatrial node. European Heart Journal. (1987);8:1249–1259. [PubMed: 3691562]

54.

Podzuweit T, Braun W, Muller A, Schaper W. Arrhythmias and infarction in the ischemic pig heart are not mediated by xanthine oxidase-derived free oxygen radicals. Basic Res Cardiol. (1987);82(5):493–505. [PubMed: 3426528]

55.

Pogwizd SM, Corr B. The contribution of nonreentrant mechanisms to malignnant ventricular arrhythmias. Basic Res Cardiol 87 Suppl. (1992);2:115–129. (Review). [PubMed: 1299206]

56.

Ravens U. Mechanoelectric feedback and arrhythmias. Prog Biophys Mol Biol. (2003);82(13):255–266. (Review). [PubMed: 12732284]

57.

Rook MB, de Jonge B, Jongsma HJ, Masson-Pevet MA. Gap junction formation and functional interaction between neonatal rat cardiocytes in culture: a correlative physiological and ultrastructural study. J Membrane Biol. (1990);118:179–192. [PubMed: 2266548]

58.

Rook MB, Jongsma HJ, deJonge B. Single channel currents of homo- and heterologous gap junctions between cardiac fibroblasts and myocytes. Pflügers Arch. (1989);414:95–98. [PubMed: 2471143]

59.

Rook MB, Jongsma HJ, van Ginneken ACG. Properties of single gap junctional channels between isolated neonatal rat heart cells. Am J Physiol. (1988);255:H770–H782. [PubMed: 2459974]

60.

Ross RS, Pham C, Shai SY, Goldhaber JI, Fenczik C, Glembotski CC, Ginsberg MH, Loftus JC. β₁ integrins participate in the hypertrophic response of rat ventricular myocytes. Circ Res. (1998);82:1160–1172. [PubMed: 9633916]

61.

Sachs F, Morris CE. Mechanosensitive ion channels in nonspecialized cells. Rev Physiol Biochem Pharmacol. (1998);132:1–77. (Review). [PubMed: 9558913]

62.

Saffitz JE, Kléber AG. Effects of mechanical forces and mediators of hypertrophy on remodeling of gap junctions in the heart. Circ Res. (2004);94:585–591. (Review). [PubMed: 15031271]

63.

Saltman AE, Aksehirli TO, Valiunas V, Gaudette GR, Matsuyama N, Brink P, Krukenkamp IB. Gap junction uncoupling protects the heart against ischemia. J Thorac Cardiovasc Surg. (2002);124:371–37619. [PubMed: 12167798]

64.

Shiraishi I, Takamatsu T, Mimikawa T, Onouchi Z, Fujita S. Quantitative histological analysis of the human sinoatrial node during growth and aging. Circulation. (1992);85:2176–2184. [PubMed: 1591834]

65.

Taggart P. Mechanoelectric feedback in the human heart. Cardiovasc Res. (1996);32(1):38–43. (Review). [PubMed: 8776401]

66.

Underwood RD, Sra J, Akhtar M. Evaluation and treatment strategies in patients at high risk of sudden death post myocardial infarction. Clin Cardiol. (1997);20:753–758. (Review). [PMC free article: PMC6655294] [PubMed: 9294665]

67.

Zeng T, Bett GCL, Sachs F. Stretch-activated whole cell currents in adult rat cardiac myocytes. Am J Physiol. (2000);278:H548–H557. [PubMed: 10666087]

68.

Zhang YH, Youm JB, Sung HK, Lee SH, Ryu SY, Ho WK, Earm YE. Stretch-activated and background non-selective cation channels in rat atrial myocytes. J Physiol (London). (2000);523(3):607–619. [PMC free article: PMC2269835] [PubMed: 10718741]