Changes in cardiac resident fibroblast physiology and phenotype in aging

Abstract

The cardiac fibroblast plays a central role in tissue homeostasis and in repair after injury. With aging, dysregulated cardiac fibroblasts have a reduced capacity to activate a canonical transforming growth factor-β-Smad pathway and differentiate poorly into contractile myofibroblasts. That results in the formation of an insufficient scar after myocardial infarction. In contrast, in the uninjured aged heart, fibroblasts are activated and acquire a profibrotic phenotype that leads to interstitial fibrosis, ventricular stiffness, and diastolic dysfunction, all conditions that may lead to heart failure. There is an apparent paradox in aging, wherein reparative fibrosis is impaired but interstitial, adverse fibrosis is augmented. This could be explained by analyzing the effectiveness of signaling pathways in resident fibroblasts from young versus aged hearts. Whereas defective signaling by transforming growth factor-β leads to insufficient scar formation by myofibroblasts, enhanced activation of the ERK1/2 pathway may be responsible for interstitial fibrosis mediated by activated fibroblasts.

Listen to this article's corresponding podcast at https://ajpheart.podbean.com/e/fibroblast-phenotypic-changes-in-the-aging-heart/.

INTRODUCTION

Cardiac fibroblasts play a role in tissue homeostasis by contributing to extracellular matrix (ECM) synthesis and deposition (via active secretion of ECM components and ECM-modifying enzymes), by modifying the phenotype of neighboring cells (via secretion of various growth factors and cytokines), and by altering the inflammatory state (via secretion of cytokines and chemokines) (46, 173). Whereas fibroblasts can change the ECM, the opposite is true as well; ECM composition and its elasticity can influence fibroblast responses to growth factors such as transforming growth factor-β (TGF-β) and promote their transition into α-smooth muscle actin (α-SMA)⁺ myofibroblasts (12). In the young heart, it is usually injuries such as myocardial infarction (MI) (32) or pressure overload (107) that induce some of the activated fibroblasts to transition into contractile myofibroblasts in response to TGF-β. The fibroblast’s phenotypic changes and its ability to mature into a functional myofibroblast determine the efficiency of healing after MI, which is compromised in the aging heart (26, 45).

Insufficient repair after MI leads to infarct expansion and adverse remodeling, which is a maladaptive matrix deposition (interstitial fibrosis) in the myocardial remote zone that ultimately leads to heart failure (79, 121). Available treatment after MI has decreased acute mortality, but the incidence of postinfarct heart failure and adverse remodeling in an aging population is increasing (141). The aged heart also manifests pathological interstitial fibrosis even without ischemia (40). Interstitial fibrosis increases the passive stiffness of the heart and contributes significantly to impairment of diastolic function (28, 144). Diastolic dysfunction limits the maximum exertion of healthy elders (180) and predisposes them to heart failure with preserved ejection fraction. There is no effective treatment for this condition, and it is now the most common type of heart failure for those over 65 yr old (175). Therefore, improving reparative fibrosis and preventing or forestalling interstitial fibrosis may reduce the incidence and severity of heart failure. The aging population that is at increased risk of diastolic dysfunction (52) and heart failure with preserved ejection fraction is projected to double in the next 40 yr.

Here, we will summarize how aging affects the cardiac fibroblast phenotype, with emphasis on the two most common cardiac pathologies, namely, replacement fibrosis (after MI) and adverse/interstitial fibrosis. Specifically, we will address and analyze the reasons for the reduced efficiency of replacement but increased adverse fibrosis in aging.

REPLACEMENT FIBROSIS: SCAR FORMATION AFTER MI

Healing After MI

The interruption of circulation that occurs during MI causes rapid cardiomyocyte death. The fetal mouse heart can regenerate after injury, but that ability is lost soon after birth (11, 136). To prevent ventricular wall rupture, in an adult heart dead cells are replaced first by a provisional matrix (fibrin clot) and then with a collagen-based scar (66). Scar formation requires tightly orchestrated processes, namely, induction of inflammation (inflammatory phase); fibroblast activation, initial scar formation, and resolution of inflammation (proliferative phase); and, finally, scar maturation (maturation phase). Although scar formation depends on various signals, the cardiac fibroblast plays a central role in promoting a scar and aiding its maturation.

Cardiac fibroblasts at the beginning of the inflammatory phase produce matrix metalloproteinases (MMPs) that degrade the ECM and ease cell migration into the infarcted area (47, 142, 159). First neutrophils and then monocytes infiltrate the damaged area in response to chemokines, some of which are produced by fibroblasts (129, 153, 172). The first wave of inflammatory cells clears the necrotic tissue and fragmented ECM. The infarcted area is then repopulated with other infiltrating immune cells [other monocytes, dendritic cells, and T cells, whose phenotype varies depending on the phase of healing (84, 148, 177)] and migrating, activated fibroblasts. Activated fibroblasts start expressing various ECM structural proteins, mostly collagen type I and type III but also other matrix proteins such as fibronectin (FN) and its splice variant extra domain-A (EDA), tenascin C, periostin, osteopontin, and thrombospondins, which regulate cellular responses and restore the scaffold destroyed after MI (32, 57, 95, 149, 187). See recent review articles detailing healing after MI (69–71, 114, 130, 137, 164).

Role of Myofibroblasts

Some activated fibroblasts mature into myofibroblasts. There are different “signaling requirements” for the activation of myofibroblast differentiation, but almost all of them involve mechanical tension achieved with increased ECM rigidity. The ECM changes activate mechanosensitive proteins via focal adhesions (FAs) (60, 87). FAs are protein complexes within which the cytoskeleton interacts with ECM (via integrin receptors) (49). Integrins, through activation of their cytoplasmic tails, bind to actin and other proteins such as talin, vinculin, and α-actinin (49, 53). The bundle of actin fibrils (including α-SMA) and associated proteins forms actin stress fibers. Formation of stress fibers is regulated by RhoA, a GTP-binding protein (39). The whole process of fibroblast-to-myofibroblast differentiation can be divided into two steps: first, fibroblasts transition into protomyofibroblasts that display strong polymerized actin fibers (F-actin) instead of depolymerized G-actin (86, 168) and, second, stress fibers contain not only F-actin but also α-SMA (88). Expression of stress fibers results in the ability of the cell to contract. This contractility is transmitted to the ECM via FAs and, in the environment of healing after MI, generates a more defined, compacted ECM (scar) and prevents infarct expansion.

As myofibroblast induction in the heart is triggered by mechanical tension, there are several pathways activated during that process. These transmit signals via transient receptor potential (TRP) channels such as TRPV4 (5), TRPM7 (59), TRPC3 (80), and TRPC6 (54). Specifically, TRPV4 and TRPC6 participate in ventricular fibroblast differentiation, and both channels are activated in the response to TGF-β, which generates Ca²⁺ influx and transcriptional upregulation of α-SMA (5, 54). Interestingly, changes in ECM stiffness (increased mechanical forces) can cause TRPV4 colocalization to FAs together with β₁-integrin (120) and its subsequent activation. Furthermore, TRPC6 has been shown to be indispensable for cardiac and dermal wound healing through the downstream activation of p38 MAPK and nuclear factor of activated T cells (NFAT) because loss of TRPC6 completely abrogates myofibroblast differentiation (54). Another pathway that is activated in response to mechanical tension and promotes myofibroblast formation is a G protein-RhoA-Rho-associated protein kinase (ROCK) pathway activating myocardin-related transcription factor (MRTF). Polymerized F-actin facilitates MRTF nuclear entry and its binding to serum response factor, a transcription factor that activates transcription of α-SMA (104, 146). Unquestionably, TGF-β plays a central role in activation of myofibroblast differentiation via a canonical Smad-dependent pathway (described below in growth factors and activation pathways affected by aging, The TGF-β Pathway) and via noncanonical pathways through TGF-β-activated kinase (TAK) and downstream either p38 MAPK, ERK1/2, or JNK. Particularly, the TGF-β-ERK-dependent activation of scleraxis and subsequent transcriptional activation of collagen type 1a2 constitute an important pathway involved in cardiac healing after MI (67, 195).

Altered Healing in the Aging Heart

Impaired healing of the aged myocardium after MI has been previously reported (26, 41). A different ECM composition and altered receptor responsiveness can contribute to this defect. In the aged heart, there is a reduced deposition of collagen in the scar and reduced expression of osteopontin and periostin (26, 191). One of the reasons for reduced deposition of collagens in the scar may be related to the elevated presence of MMP-9 in the infarct of the aged mouse heart without changes in tissue inhibitor of metalloproteinase (TIMP)-3 expression (191). Reduced TIMP-3 levels have been directly linked with elevated activity of MMP-2 and MMP-9 (167), and so the failure of TIMP-3 to increase and compensate for elevated MMP-9 may lead to collagen damage.

The transition from activated fibroblast to myofibroblast in the aging heart is defective; they display reduced chemotactic responses toward TGF-β (45), which affects their migration to where replacement fibrosis is needed. This blunted response toward TGF-β and reduced activation of the canonical TGF-β-Smad pathway (26, 45) result from reduced expression of TGF-β receptor I (TβRI) (45). Correspondingly, α-SMA expression (which is controlled by the canonical TGF-β-Smad pathway) is reduced and myofibroblast differentiation is inhibited in vitro and in the healing infarct (45). Myofibroblast differentiation can also be affected by reduced periostin expression. Periostin, a matricellular protein, in the heart is secreted by fibroblasts and stored in the ECM, but instead of playing a structural role, it rather supports ECM organization and regulates cell migration and adhesion. Periostin is often used as a marker of activated fibroblasts and promotes myofibroblast recruitment, collagen synthesis, and healing after MI (132, 152). Periostin has been shown to activate myofibroblast maturation and α-SMA expression. Fibroblasts respond to periostin via α_v-integrin (152), the expression of which has been reported to be reduced in the infarct of the aging heart (191). All that together with diminished F-actin formation and reduced activation of focal adhesion kinase (45) in aging fibroblasts suggests a wide variety of defects in this essential cell type that make the response to MI injury less effective and can lead to diastolic dysfunction with possible progression into heart failure.

INTERSTITIAL FIBROSIS IN THE ABSENCE OF INJURY

Cardiac ECM consists of a network of fibrillar proteins and proteoglycans with bound cytokines, growth factors, and ECM-modifying enzymes. The ECM not only provides a scaffold that allows structural integrity but also plays an active role in modulating cellular responses via activating integrin receptors on the cell surface. Integrins transduce signals into the cellular cytoskeleton, which further transmits signals to various pathways. These translate into migration, proliferation, differentiation, and gene expression (176).

There are two major fibrillar components of ECM: collagens and FN, both produced by fibroblasts (4, 15). Most myocardial collagen fibers consist of type I and type III. Collagen fibers connect individual myocytes and maintain myocyte alignment; therefore, changes in matrix composition can drastically affect relaxation and diastolic stiffness (98). Mass spectrophotometry analysis of matrix from young and aged left ventricles shows striking differences in various ECM components, specifically in various collagens (55). Collagen synthesis and matrix incorporation are very tightly controlled on multiple levels, as follows. Collagen peptides are first modified by prolyl and lysyl hydroxylases. Procollagen-containing COOH- and NH₂-terminal propeptides are then secreted into the extracellular space. However, to be deposited as mature fibrils, propeptides are cleaved by procollagen endopeptidases (138). Finally, collagen is stabilized by cross-linking via lysyl oxidase (158) and advance glycation end-product formation (165). Abnormally high collagen cross-linking is a primary cause of myocardial stiffness (113). Beyond the elevated deposition of collagens in the aged heart (63), increased expression of prolyl hydroxylase and lysyl oxidase has been associated with aging as well (143, 161), perhaps further promoting ECM stiffness. Secreted protein acidic and rich in cysteine (SPARC), a matricellular protein, is also upregulated in the aging heart, and it has been shown to influence procollagen processing by cardiac fibroblasts and promote mature collagen deposition (81). Increased expression of periostin by aging cardiac fibroblasts has been previously reported (110). All these may contribute to increased tissue stiffness, which correlates with elevated ECM component synthesis and deposition and fibroblast phenotypic changes (12, 91, 117). ECM mechanosensing is transmitted through integrins. Differences in the expression of various integrins in the young and aged heart have been reported by others and discussed below (123). Transcriptional cofactors such as Yes-associated protein (YAP) and transcriptional coactivator with PDZ-binding motif (TAZ) have been recently discovered as mechanosensors (62); they interact with Smads and other transcription factors to switch the cells toward a profibrotic phenotype (9, 118, 133, 163). YAP/TAZ levels or activation in the aging cardiac fibroblasts are unknown, but fibroblasts isolated from aging skeletal muscle display increased nuclear levels of both mechanosensors (160) consistent with the notion that the injured aging muscle acquires a profibrotic phenotype (93).

Cardiac fibroblasts express not only ECM components but also ECM-modifying enzymes such as MMPs, a disintegrin and metalloproteinases (ADAMs), ADAMs with thrombospondin repeats, and TIMPs [for a review, see Spinale (159)]. Whereas proteases degrade the matrix and promote remodeling, they also generate discrete fragments of various ECM components (collagen and FN) that have bioactive properties (75, 169, 190). Proteases also liberate growth factors from their binding proteins (68, 194) or latency-associated protein (194).

FN, the other essential component of ECM, plays a role in the regulation of cellular processes and also serves as a scaffold protein, directing the assembly of other ECM components (156). FN is a multidomain glycoprotein composed of independently folded modules termed types I, II, and III (140). An EDA-containing splice variant of FN is normally made only during fetal life and not by adult fibroblasts. However, during injury, adult fibroblasts synthesize EDA (149). Chronic ECM assembly with the presence of EDA switches normal tissue repair to dysregulated fibrosis (21), which has been observed in the aged heart (44). FN molecules are organized into fibrils of various thicknesses, whereby dimers or multimers form high-molecular weight FN (so-called “super-FN”) (131). Cross-linking and multimerization of FN are enhanced in the presence of an FN fragment, a peptide named “anastellin” (127, 131); however, it is not known whether anastellin expression is increased in the aged heart.

All these ECM proteins in the young heart are synthesized by fibroblasts in response to activation of the canonical TGF-β pathway (89, 99, 100), which seems to be only partially functional in the aging fibroblast but which may be controlled by other pathways as well (see below).

GROWTH FACTORS AND ACTIVATION PATHWAYS AFFECTED BY AGING

The TGF-β Pathway

TGF-β is a cytokine that activates fibroblasts via binding to its receptor II (TβRII), causing recruitment of TβRI and its autophosphorylation. There are two TβRIs, activin receptor-like kinase (Alk)5 and Alk1. Alk5 is the main TβRI expressed in fibroblasts, but there are reports that have demonstrated that Alk1 can also be operant in fibroblasts in response to TGF-β (128). Activated Alk5 recruits and phosphorylates Smad proteins (Smad2 and Smad3 for the canonical pathway). Phosphorylated Smads in the cytoplasm interact with Smad4, which facilities their nuclear transportation. Within the nucleus, Smads engage in transcriptional regulation of various genes (25). ECM components such as collagen, FN, vimentin, and various MMPs that promote remodeling and α-SMA-dependent differentiation into myofibroblasts are controlled by the TGF-β pathway (56, 64, 115, 122, 189). That the TGF-β signaling pathway is essential for synthesis/processing of ECM proteins is supported by the following evidence: 1) cardiac TGF-β overexpression in young mice causes cardiac fibrosis (2), 2) fibroblast-specific deletion of TβRI/II or Smad2/3 attenuates fibrosis in a model of pressure overload in young mice (99), and 3) aged TGF-β heterozygous mice show reduced cardiac fibrosis (24). Paradoxically, fibroblasts derived from the aged mouse heart have reduced expression of TβRI [Alk5; ~50% of the levels observed in young fibroblasts (45)] and display reduced activation of the Smad2/3 pathway in response to TGF-β (26, 45). It has been previously demonstrated that prolonged TGF-β stimulation downregulates expression of both TβRs (16), and we hypothesize that TβRI downregulation may be related to increased availability of active TGF-β in the aged heart. TGF-β secretion by fibroblasts derived from young and aged mouse hearts seems to be unchanged (K. A. Cieslik, unpublished data); however, TGF-β mRNA expression in the aged heart is elevated (43), perhaps because of the higher number of infiltrating macrophages and T cells that secrete TGF-β as well. TGF-β is stored in the ECM in a latent form not accessible to its receptors (7). It is bound in a complex with latency-activated peptide and latent TGF-β-binding protein. Release of TGF-β from the ECM is mediated via several mechanisms (90), involving MMPs or integrins. Activation of TGF-β can be achieved via proteolitic cleavage by MMP-2 and MMP-9 (194). MMP-9 (but not MMP-2) expression is upregulated in the uninjured left ventricle in the aged mouse heart (36) concomitantly with reduced presence of TIMP-3 (20), suggesting that there may be increased release of TGF-β from the matrix in the aging heart by this mechanism. MMP-9 deletion attenuates cardiac fibrosis and diastolic dysfunction in aging mice (37). TGF-β activation via integrins in the heart involves α_vβ₃-integrin and α_vβ₅-integrin binding to latency-activated peptide and via integrin cell-mediated force liberating TGF-β from the complex (147). Interestingly, β₃-integrin knockout mice are fairly protected from cardiac fibrosis in a model of pressure overload (13), concomitant with the role of α_vβ₃-integrin in TGF-β release from the matrix. In contrast, β₁-integrins may have an opposing role: β₁-integrin levels are reduced in aged murine hearts (27, 116) and in aged myocytes (48), and cardiac myocyte-specific reduction of β₁-integrin results in fibrosis (150).

Aberrant TGF-β signaling in aging has been characterized not only in cardiac fibroblasts by our laboratory (45) and others (26) but also in other cell types such as chondrocytes (19), macrophages (134), skin fibroblasts (83), and neurons (166), all leading to reduced activation of the canonical Smad2/3 pathway. Recently, it has been demonstrated that in aging cartilage, the other TβRI (Alk1) and its downstream effectors (Smad1/5/8) become predominantly engaged in TGF-β signaling, causing phenotypic changes in chondrocytes (178). The TGF-β-Alk1-Smad1 pathway may be triggered by elevated expression of endoglin, a TGF-β type III receptor (124). Reduction of endoglin expression decreases TGF-β-activated collagen type I levels and fibrosis and improves cardiac function (96). Altered expression of endoglin in aging has been reported in the myeloid and endothelial lineages (8, 18).

The aging human and mouse hearts are characterized by increased ECM production leading to interstitial fibrosis (42, 170, 171). Most of the collagens that comprise the bulk of ECM proteins are synthesized by fibroblasts. Because of apparent downregulation of the canonical TGF-β signaling pathway in aging fibroblasts, cardiac fibrosis in aging may result from TGF-β independently or the synergistic interaction of residual TGF-β signaling plus that of other signaling pathway(s). A notable possibility for the involvement of another pathway is that of ERK1/2 activation.

The ERK1/2 Pathway in Interstitial Fibrosis

Several other pathways are directly or indirectly activated by TGF-β. In one of these noncanonical TGF-β pathways, activation of TβRs phosphorylates ShcA, which then recruits the growth factor receptor-bound protein-2 (Grb2)/son of sevenless (Sos) complex and activates the Ras-Raf-Mek1-ERK1/2 pathway downstream (106). However, it was also shown that pharmacological inhibition of TβRI kinase activity does not fully suppress ERK activation, suggesting that other signaling pathways are involved (92). There are different potential ligands that can activate downstream ERK signaling in cardiac fibroblasts, PDGF and EGF among many. It has been previously demonstrated that PDGF-α, PDGF-β, and EGF can induce fibrosis (74, 76, 119, 157) and activate the ERK pathway (139). Whereas the canonical TGF-β-Smad-dependent signaling in fibroblasts derived from aged hearts was severely reduced, we identified that the Ras-ERK pathway was pathologically activated and drove collagen overexpression (42). Furthermore, fibroblasts derived from the aged mouse heart respond to pathophysiological levels of insulin by upregulating an already elevated ERK phosphorylation and increasing collagen type I expression even further (42). Insulin-dependent ERK activation has also been reported by others (101, 103). It is not known whether there is cross talk between any of these pathways (TGF-β, PDGF, EGF, and insulin) in aging cardiac fibroblasts, although weak synergism between insulin and EGF has been reported in other models (22).

Another reason for the elevation of ERK activation (other than upregulation of stimulating signals) in aging may involve downregulation of negatively controlling signals; protein phosphatase 2A (106, 155) and dual-specificity phosphatases (DUSPs) (34) negatively regulate ERK activity by removing the phosphate moiety. It is not known whether aging exerts any changes in activity of ERK-regulating phosphatases, but the expression levels of DUSP1 and protein phosphatase 2A are severely downregulated by 12- and 5-fold, respectively (185).

ERK activity can also be increased by reactive oxygen species (1). Much higher levels of superoxide anion were produced by fibroblasts derived from aged hearts than those from young hearts (43, 171). It has also been reported that the reactive oxygen species pathway activates expression of various cytokines and collagen (112) and the aging fibroblast expresses higher levels of IL-6, monocyte chemoattractant protein (MCP)-1, and other cytokines (43, 46). Fibroblasts acquire a proinflammatory phenotype while activated during ischemic insult (154) and then transition into a profibrotic state. Interestingly, in aging in the steady state, this proinflammatory but also profibrotic phenotype is found to be increased in vivo and in vitro (42, 46), suggesting that the switch-off mechanism is not functioning. In the young heart, activated fibroblasts undergo deactivation or apoptosis (58, 183, 196), but in the aged heart the fate/lifespan of these cells is unknown. However, persistent activation of these fibroblasts has been noted (42, 171).

Other Possible Pathways Involved in Fibroblast Activation That May Be Affected by Aging

Angiotensin II-dependent fibroblast activation.

Angiotensin II (ANG II) signals via activation of the ANG II type 1 receptor (AT₁R), which has been identified on cardiac fibroblasts (182). It mediates proliferation and activation of fibroblasts and increases collagen expression (35, 77). Both ANG II and AT₁R levels are elevated in the aging heart (31, 40, 51, 85) and have been involved in fibrosis (197). ANG II is not only increased via systemic elevation but also produced locally in the heart (51). ANG II signaling also cross-talks with the TGF-β and ERK1/2 pathways (30, 109, 184) and also the p38 MAPK pathway, causing periostin upregulation (109).

The p38 MAPK pathway.

The cardiac fibroblast-specific p38 MAPK pathway can be activated via a noncanonical TGF-β-TAK1 pathway (45) or via TRPC6 (54). The TGF-β-TAK1-p38 MAPK pathway controlling myofibroblast maturation in aging is impaired because of reduced expression of the Alk5 receptor (45), and the pharmacological activation of p38 MAPK downstream of Alk5 rescues aged myofibroblast maturation (45). The influence of aging on the expression level of TRPC6 in cardiac fibroblasts is unknown; however, the levels are elevated in aging aortas (65). It has also been demonstrated that DUSP1, a phosphatase that restrains (together with DUSP4) p38 MAPK activation (10), is severely downregulated in the aging heart (185).

Activation of adrenergic receptors.

β₂-Adrenergic receptors (β₂-ARs) are expressed by cardiac fibroblasts and play a significant role in adrenergic-dependent downregulation of collagen synthesis and myofibroblast formation (162) via cAMP signaling. However, chronic overstimulation of β₂-ARs causes proliferation, migration, myofibroblast differentiation, and collagen secretion (174). This can happen when G protein βγ-subunits recruit upregulated G protein-coupled receptor kinase 2 (GRK2), which engages β-arrestin (33). The GRK2-β-arrestin pathway suppresses β₂-AR signaling and promotes a profibrotic phenotype via ERK and Smad pathways (94, 108). β-Adrenergic stimulation has been shown to be blunted in fibroblasts derived from aged rat hearts (38), and β-arrestin expression was significantly increased in cardiac fibroblasts isolated from human failing hearts (108) consistent with elevated profibrotic signals and increased ECM deposition.

POTENTIAL THERAPEUTIC TARGETS

We have analyzed the existing literature to delineate the differences in phenotype and physiology of cardiac fibroblasts derived from young and old hearts. Current proposed therapeutic targets are based on experimental models that mostly use young animals, which may not be applicable to treating fibrosis (either reparative or adverse) in the aged heart. For example, the unquestioned role of TGF-β in the fibrosis in young heart injury models becomes problematic with the aged heart in mind. Will further blockade/reduction of the defective TGF-β pathway in old hearts be beneficial or even more detrimental? It is a challenging question since there is an apparent downregulation of the canonical TGF-β pathway in aging, but, on the other hand, it has been demonstrated that reduction of TGF-β (as in the TGF-β heterozygous mouse model) reduces cardiac fibrosis in aged mice (24).

While analyzing activation pathways that are upregulated in the aging fibroblast, we note that the ERK pathway may play a central role in promoting the profibrotic phenotype. Moreover, ERK seems to be at the intersection between various signaling pathways; it can be activated by TGF-β, ANG II, PDGF (via increased receptor expression), or adrenergic receptor uncoupling (via β-arrestin), all of which are upregulated in aging heart fibroblasts (38, 43, 51, 97, 108). This suggests that blocking these pathways can be potentially beneficial for the aging heart. AT₁R blockers (in patients) (151) and G protein βγ-subunit-GRK2 signaling inhibitor (33) or an inhibitor of the PDGF-mediated pathway (111) in experimental models improve heart function and reduce fibrosis.

Another possible venue to explore is the role of inflammation in activation of the profibrotic phenotype. Aging fibroblasts express various cytokines at higher levels compared with young controls (46). Elevated levels of MCP-1 attract leukocytes from the blood (72), and IL-6 promotes monocyte maturation into collagen-expressing macrophages (46). Interestingly, reduction of leukocyte infiltration protects the older heart from fibrosis and prevents activation of fibroblasts (171), suggesting that there is an active feedforward loop between resident fibroblasts and infiltrating leukocytes. The expression of two major cytokines by aged fibroblasts (MCP-1 and IL-6) is controlled by the above-mentioned ERK pathway as well (46).

This exaggerated induction of the proinflammatory phenotype, which is a feature of the senescence-associated secretory phenotype, can also cause disruption of homeostasis of other cells such as cardiomyocytes. It has been demonstrated that fibroblast-secreted factors can alter the cardiomyocyte’s electrophysiological state, reduce its contractile force, and promote hypertrophy (105, 181). That implies that agents that reduce inflammation can be beneficial for the aging heart.

FIBROBLAST DIVERSITY

Fibroblasts of mesenchymal origin are quite abundant in the heart. Depending on the criteria used to identify fibroblasts, the mouse heart comprises ~20% of fibroblasts (14, 135), whereas the human heart has ~40% of mesenchymal cells including fibroblasts (17).

Fibroblasts can assume various phenotypes depending on age (45) or pathophysiological state (resting vs. activated) (188, 192). Their functional/phenotypical heterogeneity, even within the same organ [atria vs. ventricles (193) or left vs. right ventricles (78)], may be due to their various developmental sources. In the adult heart, in addition to activation of the resident fibroblast pool, fibroblasts may derive from the endothelial lineage (126), epicardial population (6, 126, 179), bone marrow (82, 125), pericytes (61), and mesenchymal cells of perivascular origin (102). The functional defects described above may also arise from partial depletion of specific specialized progenitors. We have shown that mesenchymal stem cells (MSCs) that give rise to most of the resident fibroblasts in the heart are dysfunctional with aging and have reduced expression of stemness markers (45) partially caused by the downregulation of canonical TGF-β signaling (145). Because of reduced functionality of the TGF-β pathway in cardiac MSCs (45), it may permit not only MSC differentiation into dysfunctional fibroblasts but also fibroblast excessive proliferation leading to interstitial fibrosis. It has been shown that in the aging muscle (in vivo) and heart (in vitro), progenitors adopt an adipocyte/fibroblast lineage (45, 93). This may suggest that these particular bipotential progenitors are preferentially activated or proliferating in aging.

Recently, several genetically modified animals have been created to investigate fibroblast origins in the embryonic and adult mouse heart. Via epithelial-to-mesenchymal transition, the epicardium contributes to the fibroblast pool in the embryonic heart. Tracing of embryonic fibroblasts can be achieved by Cre mouse lines under control of T-box transcription factor 18, Wilm’s tumor protein (29, 186), or transcription factor 21 (3) or using a reporter line labeling PDGF receptor-α-expressing fibroblasts (157). Adult fibroblasts can be traced using PDGF receptor-α or collagen type 1a (as reporter lines) (126, 135) or using Cre indicator lines such as transcription factor 21 or periostin (73, 95). Fibroblasts can be traced not only in the steady state (135) but also during MI/post-MI (73, 95) or after pressure overload-induced fibrosis (126). However, none of these mouse lines have been studied in aging.

SUMMARY

We discussed here two pathophysiological states: MI and adverse fibrosis. Inadequate healing and poor scar formation in the aging heart can lead to adverse remodeling. At the same time, fibrosis can occur even in the uninjured myocardium in older patients. Both these pathologies can lead to heart failure. Approximately 5.1 million patients in the United States alone have been diagnosed with heart failure (23, 175), and 80% of these patients are 65 yr or older (175). With an extended lifespan and aging worldwide population, it is necessary to broaden our understanding of the biology of aging, perhaps even to as specific a consideration as changes in an individual cell type.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grant R01-HL-089792, the Medallion Foundation, and the Hankamer Foundation.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

All authors contributed equally to all aspects of the study from conception to publication.