The Origin and Arrhythmogenic Potential of Fibroblasts in Cardiac Disease

Abstract

Fibroblasts play a major role in normal cardiac physiology and in the response of the heart to injury and disease. Cardiac electrophysiological research has primarily focused on the mechanisms of remodeling that accompany cardiac disease with an emphasis on myocyte electrophysiology. Recently, there has been increasing interest in the potential role of fibroblasts in cardiac electrophysiology. This review focuses on the arrhythmia mechanisms involving interactions between myocytes and fibroblasts. We also discuss the available evidence supporting the contribution of intracardiac and extracardiac sources to the fibroblast and myofibroblast populations in diseased hearts.

Introduction

Cardiovascular disease remains a major cause of death in the United States. The lifetime risk of developing coronary artery disease after the age of 40 is 49 % for men and 32 % for women [1]. Most of these deaths are associated with the development of malignant ventricular arrhythmias, yet the underlying mechanisms responsible for initiation and maintenance of ventricular arrhythmias remain poorly understood. Cardiac electrophysiological research has primarily focused on the mechanisms of remodeling that accompany cardiac disease with an emphasis on myocyte electrophysiology. Recently, there has been increasing interest in the potential contribution of the non-excitable cell populations to cardiac electrophysiology and arrhythmia mechanisms [2–17]. Fibroblasts are the major non-myocyte cell type and contribute to the structural, mechanical, and electrophysiological properties of the heart [18–20]. Fibrosis is present in a variety of cardiac diseases associated with a high incidence of arrhythmias. The predominant cellular mechanism of fibrosis involves the emergence of activated fibroblasts or myofibroblasts. Traditionally thought to originate from proliferation and activation of resident fibroblasts, recent studies have demonstrated myofibroblasts can also originate from other intracardiac and extracardiac sources [21–42]. Fibroblasts are known to be a heterogeneous population of cells, and it is currently unknown if phenotypic differences are associated with these different origins [43–50]. This review focuses on the contribution of fibroblasts to cardiac electrophysiology and arrhythmia mechanisms. We also discuss available evidence supporting the contribution of resident and non-resident sources to the fibroblast and myofibroblast populations in diseased hearts.

Cardiac Fibroblast Functions

Fibroblasts are highly responsive to the chemical and mechanical environment of the heart. Figure 1 summarizes the primary factors affecting fibroblasts and their major functional responses. The primary function of fibroblasts in the normal heart is maintenance of the extracellular matrix which requires tight regulation of synthesis and degradation pathways [51]. Fibroblasts synthesize fibrillar collagen as a precursor polypeptide which is further processed after cellular export and crosslinked to form mature collagen. Fibroblast collagen synthesis is transcriptionally regulated by fibrogenic growth factors including transforming growth factor β (TGFβ) [51]. Fibroblasts also coordinate degradation of collagen through secretion of matrix metalloproteinases and tissue inhibitors of matrix metalloproteinases.

Fig. 1.

Functions of cardiac fibroblasts. Fibroblast function is affected by a number of stimuli including chemical mediators, neurotransmitters, mechanical forces, and injury. Fibroblasts respond to stimuli by releasing chemical mediators and altering the synthesis and degradation of extracellular matrix (ECM) proteins, ATP release, connexin expression, migration, proliferation, and adhesion properties

Cardiac myocytes have limited regenerative capabilities in response to cardiac disease and injury as dead and apoptotic myocytes are replaced with fibrotic tissue. Fibroblasts play a major role in the response to cardiac injury and the development of fibrosis. Critical to the injury response is the emergence of activated fibroblasts or myofibroblasts which are not present in the normal cardiac muscle. The emergence of myofibroblasts is strongly promoted by TGF-β1 [52]. Myofibroblasts have characteristics that are intermediate between fibroblasts and smooth muscle cells, express α-smooth muscle actin (α-SMA) and contractile proteins, and have higher rates of cellular proliferation and extracellular matrix deposition compared to fibroblasts [53–55]. The ultrastructural features of myofibroblasts include myofilaments, well-developed rough endoplasmic reticulum, and extensive cell–matrix contacts [56–58]. During wound healing, myofibroblasts strengthen injured tissue by providing additional extracellular collagen.

Cardiac tissue repair in response to myocardial infarction is a dynamic process and includes homeostasis, infiltration of immune and inflammatory cells, degradation and phagocytosis of dying cells and debris, repopulation of the zone of injury with fibroblasts and myofibroblasts, extracellular matrix remodeling, and formation of a mature scar [51, 59]. Fibroblasts and myofibroblasts are essential during all phases of tissue repair. During the initial injury response, fibroblasts and myofibroblasts secrete matrix metalloproteinases and other cytokines to facilitate recruitment of other fibroblasts to the injured region and facilitate the degradation of extracellular matrix. Recent studies have demonstrated that cardiac fibroblasts release ATP into the extracellular space through connexin hemichannels that could further activates profibrotic cellular responses through the MAPK pathway [60]. Following the initial injury response phase, fibroblasts infiltrate the injured region and secrete extracellular matrix proteins to maintain the structural integrity of the affected area and facilitate the formation of scar tissue. Termination of the injury response occurs following the establishment of a mature scar through apoptosis of activated fibroblasts. Unlike the injury response in other tissues, a small number of cardiac myofibroblasts are still present in the scar region several years after injury [61–63].

Origin of Fibroblasts

During heart development, cardiac fibroblasts originate from mesenchymal cells which are principally derived from the embryonic epicardium [64–68]. Other sources of fibroblasts have been proposed including endocardial cushion, epithelial-to-mesenchymal transformation [69], and postnatal recruitment of circulating bone marrow cells [70]. Fibroblasts and myofibroblasts in injured hearts originate from intracardiac and extracardiac sources including expansion and activation of resident fibroblasts, epicardial cells that undergo epithelial-to-mesenchymal transition, endothelial cells that undergo transformation to mesenchymal cells, and circulating bone marrow-derived progenitor cells [24, 25, 27, 28, 30, 31, 40, 71–73] (see Fig. 2).

Fig. 2.

Cellular sources that contribute to the cardiac fibroblast population following injury. In response to cardiac injury, resident cardiac fibroblasts proliferate and migrate to the site of injury. Fibroblasts in injured sites are also derived from transformed epicardial, endothelial, and bone marrow sources

Intracardiac Sources of Fibroblasts

Resident Fibroblasts

It has been widely believed that a significant number of myofibroblasts present in fibrotic hearts are derived from the resident fibroblast population. Initial studies investigating the contribution of non-resident fibroblast sources supported this idea [22]. However, recently, several studies have challenged this concept by demonstrating that multiple cellular lineages contribute to a significant percentage of the cardiac fibroblast population under pathological conditions [24, 25, 27, 28, 30, 31, 40, 71–73]. The ability to determine the fate and contribution of resident cardiac fibroblasts under pathological conditions has been limited by the lack of specific markers for resident fibroblasts. Evidence supporting resident fibroblast contribution to the cardiac fibroblast population under pathological conditions has been obtained from studies of chronic cardiac transplant rejection [26, 29]. In a rat heart transplantation model, fibroblasts from allografts undergoing chronic rejection were shown to be derived from recipient (extracardiac) and donor (intracardiac) sources [29]. More recently, studies using endomyocardial biopsy samples from areas of increased cardiac fibrosis in heart transplant patients have indicated that fibroblasts are mainly derived from intracardiac sources [26]. These studies did not differentiate between fibroblast and myofibroblast populations. There were several important differences between these studies including the use of immunosuppression drugs in human patients and the use of antibodies specific to macrophages to better differentiate those cells from fibroblasts. Importantly, these studies did not exclude the possibility that a subpopulation of cardiac fibroblasts may have originated from the intracardiac sources discussed below.

Epithelial-to-Mesenchymal Transition-Derived Fibroblasts

The epicardium is the outermost layer of the heart and is composed of epithelial cells. During development, epicardial cells undergo epithelial-to-mesenchymal transition and give rise to multiple cell types including fibroblasts, smooth muscle cells, and endothelial cells. In the adult normal heart, epicardial cells are quiescent and do not undergo epithelial-to-mesenchymal transformation [41, 42]. However, recent studies have indicated that the epicardium is responsive to cardiac injury and plays an important role in cardiac repair [35, 39, 41, 42]. In vitro studies have shown that human epicardial cells spontaneously undergo epithelial-to-mesenchymal transition and the antigen profile of the transformed cells resembles subepithelial fibroblasts [39]. More recently, studies using animal models of myocardial infarction have shown that cardiac injury leads to epicardial cell proliferation and stimulates the formation of epicardium derived cells [35, 41, 42]. Epicardium-derived cells form a thick layer on the surface of the heart over the injured region and adopt fibroblast, myofibroblast, and smooth muscle cell phenotypes. The epicardial expansion and epithelial-to-mesenchymal transition have been shown to be dependent on Wnt1 signaling [35]. Additional studies are needed to determine whether Wnt1 signaling differentially affects the epithelial-to-mesenchymal transition-derived fibroblast and myofibroblast populations.

Endothelial-to-Mesenchymal Transition-Derived Fibroblasts

Endothelial cells line the interior surface of blood vessels and are in direct contact with the circulating cells of the blood. During development, endothelial cells undergo endothelial-to-mesenchymal transition and give rise to the atrioventricular cushion, the primordial of the valves and septa [74]. In the normal adult heart, endothelial cells do not significantly contribute to the cardiac fibroblast population [32, 33]. However, in response to cardiac disease or injury, endothelial cells undergo endothelial-to-mesenchymal transition and give rise to fibrosis and the accumulation of a significant number of fibroblasts and myofibroblasts [31, 32, 37, 40].

The percentage of total fibroblasts derived from endothelial sources may be dependent on the pathological stimulus. In a pressure overload model of heart failure, between 27 % and 35 % of the total fibroblast population is derived from endothelial cells [31]. A smaller percentage of fibroblasts of endothelial origin, 15 % to 20 % of all fibroblasts, were reported in a streptozotocin-induced diabetes model [40]. Differences in the number of myofibroblast derived from endothelial sources have also been reported. In pressure-overloaded hearts, approximately 75 % of myofibroblasts have endothelial origin [31]. In animal models of myocardial infarction, between 35 % and 40 % of α-SMA positive cells, some of which are likely to be myofibroblasts, are derived from endothelial cells [32].

Endothelial-to-mesenchymal transition to fibroblasts and myofibroblasts has been shown to be induced by TGF-β1 in a Smad3-dependent manner and inhibited by bone morphogenetic protein 7 and hepatocyte growth factor [31, 37]. Endothelin-1 can also stimulate endothelial cells to undergo endothelial-to-mesenchymal transition through inhibition of TGF-β1 signaling [40]. Activation of Wnt signaling causes endothelial cells to undergo endothelial-to-mesenchymal transformation, and the transformed cells can be identified by Wnt activity [32].

Extracardiac Sources of Fibroblasts

Bone Marrow-Derived Fibroblasts

Bone marrow-derived cells are recruited to the heart following cardiac injury and play a critical role in inflammation and cardiac repair [23–25, 27, 28, 30]. Several studies have demonstrated that a significant number of fibroblast and myofibroblasts in the infarcted myocardium are derived from bone marrow cells [23–25, 27, 28, 30]. In animal models of myocardial infarction, bone marrow-derived fibroblasts and myofibroblasts densities peak between 1 to 3 weeks after injury. During this period, bone marrow-derived fibroblasts represent between 25 % and 30 % of the fibroblast population, while bone marrow-derived myofibroblasts represent between 24 % and 57 % of the myofibroblast population in the injured region [24, 25, 28]. Studies have shown that the percent of fibroblast that are bone marrow-derived remains relatively constant for up to 4 months [25]. On the other hand, the percent of bone marrow-derived myofibroblasts decreases after 3 weeks and remains constant for at least 4 months [24, 25]. Bone marrow-derived fibroblasts and myofibroblasts are absent from areas remote from the infarct or perinfarcted regions [24, 28]. Recent studies have demonstrated that bone marrow-derived cells are actively replaced in the infarcted heart, and this process continues for several months after injury [27]. Specifically, bone marrow-derived fibroblasts in the injured region are turned over within 2 weeks after infarction [27]. Bone marrow-derived myofibroblasts participate actively in scar formation by producing collagen I in the injured region [28]. Significant numbers of bone marrow-derived fibroblasts and myofibroblasts have also been identified in senescent animals and other models of cardiac disease [23, 31, 34, 38].

Arrhythmogenic Potential of Cardiac Fibroblasts

Fibrosis leads to changes in the mechanical properties of the heart, disrupts electrical connectivity between myocytes, and impairs myocyte oxygen availability. Fibrosis and interactions between fibroblasts and myocytes can alter cardiac electrophysiology and contribute to arrhythmia formation through a number of mechanisms illustrated in Fig. 3. Traditionally, fibroblasts and fibrosis are considered to alter cardiac electrophysiology by mechanically separating myocytes and creating barriers to propagation of the electrical impulse (Fig. 3a) [75–77]. Such conduction disturbances can produce conduction delay or block that leads to the formation of functional reentry.

Fig. 3.

Arrhythmogenic mechanisms of fibroblasts myocyte interactions. a Fibrotic regions act as physical or functional barriers to electrical conduction. Electrical wavefronts that encounter such regions break at the proximal side and reconnect on the distal side. b Fibroblast–myocyte electrical coupling allows for slow electrotonic conduction through fibrotic regions. c Partial electrical isolation of myocytes can lead to successful electrical conduction of impulses originating from ectopic foci. d Chemical mediators released from fibroblasts can exert paracrine effects on myocytes leading to slow conduction

There is increasing evidence that fibroblasts may also actively contribute to cardiac electrophysiology and arrhythmogenesis through direct electrical coupling and paracrine mechanisms [2–17]. Fibroblasts are electrically passive cells that do not express sodium or other excitatory currents [50]. In addition, fibroblasts have high input impedances and more depolarized resting membrane potentials compared to cardiac myocytes [50]. In vitro and in vivo evidence suggests that fibroblasts and myocytes can electrically couple and that fibroblasts can affect myocyte resting potential and action potential parameters [6–9, 78, 79]. Together, these findings indicate that fibroblasts can directly modulate cardiac electrophysiology and surrounding myocytes through electrotonic mechanisms. These interactions can result in a number of conduction disturbances. Myocyte–fibroblast interactions facilitate conduction by allowing electrotonic propagation of the electrical activation between physically separated myocytes (Fig. 3b). Increasing densities of fibroblasts favor conduction slowing and eventually conduction block [6–8]. Recent studies have suggested that pharmacological disruption of the actin cytoskeleton can decrease these effects [3]. Fibroblasts can also contribute to arrhythmia initiation by influencing automaticity [5, 9, 80]. Partial electrical isolation of spontaneously active cells by coupling to fibroblasts can facilitate successful conduction from ectopic foci (Fig. 3c). Recent studies have also demonstrated that fibroblast–myocyte interactions can result in depolarization-induced automaticity in surrounding myocytes. Fibroblast can also increase automaticity by increasing the effect of ionic and oxidative stress on myocytes [12].

Studies have shown that fibroblasts can also alter myocyte electrophysiological properties through paracrine mechanisms (Fig. 3d). It is well-established cardiac fibroblasts release factors that contribute to pathological remodeling, and it is becoming increasing clear that these factors can also contribute to arrhythmogenesis [10, 51, 81, 82]. Paracrine factors released by cardiac fibroblasts have been shown to modulate myocyte action potential parameters and conduction properties [8, 10]. In addition, studies from our laboratory have demonstrated that the paracrine effects of fibroblasts on myocyte electrophysiology are enhanced following cardiac injury [8]. Fibroblast paracrine effects are dose dependent and partially reversible. Gene expression studies have indicated that these changes are the result of a reduced expression of fast sodium current, inward rectifier current, and the transient outward potassium current in myocytes treated with fibroblast paracrine factors [10]. Myocyte connexin43 (Cx43) expression, phosphorylation, and function are unaffected by fibroblast paracrine factors [10].

Traditionally, fibroblasts have been considered to be a homogenous population of cells. However, recent data support the idea that fibroblasts are phenotypically distinct depending on developmental stage, organ, and physiological conditions [8, 43–50]. This heterogeneity, particularly potential differences in ionic currents and connexin expression, has important implications with regards to modulation of the arrhythmogenic substrate. Studies of fibroblast connexin expression during pathological conditions have shown that fibroblasts in ventricular infarct scar tissue express Cx43 or connexin45 (Cx45) with spatially and temporally distinct patterns [83]. Connexin 40 (Cx40) has not been identified in these cells. Fibroblasts expressing Cx45 infiltrate damaged tissue within the first few hours after infarction, reach their peak density within 6 days, and decrease thereafter. The number of Cx43 expressing fibroblasts starts increasing 6 days after infarction and continues to rise until at least the fourth week. These data suggest that Cx45 may be responsible for electrical coupling between fibroblasts and myocytes during the acute remodeling process, while Cx43 may be involved at later stages. It is currently unknown whether these cells represent a single fibroblast population that initially expresses Cx45 and then Cx43, or if there are multiple fibroblast populations involved. The latter is supported by evidence showing that Cx45 and Cx43 do not colocalize. Recent studies have demonstrated that connexin levels and coupling to myocytes are elevated following cardiac injury which increases the potential of fibroblasts to influence myocyte electrophysiology [8, 84]. It is intriguing to speculate that the populations may correspond to quiescent or activated fibroblasts with different lineages. A better understanding of the origin and function of the different fibroblast populations may provide valuable insight into possible arrhythmogenic mechanisms and therapeutic approaches.

Conclusions

In this review we have discussed the role of cardiac fibroblasts in healthy and diseased hearts and described how fibroblasts contribute to the arrhythmogenic substrate. We have also reviewed the available evidence demonstrating fibroblasts and myofibroblasts originate from a number of lineages in diseased hearts. Fibroblasts and myofibroblasts originate from epithelial-to-mesenchymal transition, endothelial-to-mesenchymal transition, and bone marrow-derived sources. Due to the lack of specific markers, the relative contribution of resident fibroblasts in cardiac injury and disease is currently unclear. Studies have also indicated the relative contribution of these populations changes with time after injury and possibly with pathological stimuli. Further experimentation is needed to determine whether populations of fibroblasts derived from different sources are phenotypically distinct and the potential functional roles of each of these populations in cardiac electrophysiology. These studies would have wide-ranging implications for the treatment of cardiac arrhythmias and may significantly contribute to our understanding of the basic principles that govern electrophysiology in healthy and injured hearts.