Signals from the extracellular matrix: Region- and sex-specificity in cardiac aging

Abstract

During aging, the cardiac extracellular matrix (ECM) undergoes gradual remodeling that reduces the heart’s ability to function. Specific ECM changes cause alterations in cellular signaling pathways, eliciting maladaptive responses. Here, we provide insight into the current knowledge of how age-specific ECM changes contribute to altered ligand–receptor interactions, dysregulated mechanotransduction, and the propagation of pro-fibrotic signaling cascades that underpin dysfunction. We also highlight regional and sex differences that new biomolecular and bioengineered technologies have recently uncovered. We call for new biomaterial strategies that mimic spatiotemporal and sex-specific ECM alterations to equip researchers with the tools to unravel complex cellular signaling events. We believe this can be achieved through interdisciplinary cooperation amongst researchers spanning matrix biology, biomaterials, spatial omics, and biomedical engineering.

Introduction

Recent multi-omics studies have revealed the nonlinear progression of age-related molecular dysregulation in humans, with major changes in pathways related to cardiovascular health occurring around the age 60 [1]. Cardiac aging is associated with a decline in heart function, including hypertrophy, reduced cardiac output, and diastolic dysfunction [2]. During healthy aging, heart tissue stiffness gradually increases, extracellular matrix (ECM) accumulates, and ECM architecture becomes disrupted [3]. Matrix mechanical cues (e.g., increase in tissue or fiber mechanics) are transmitted to cells via mechanotransduction. Changes in ECM composition (e.g., higher ECM abundance or matrix-immobilized factors) affect cell–matrix interactions [4,5]. Re-organization of ECM (e.g., fiber disarray) leads to topographic cues that dictate cellular alignment, which are especially important in the heart [6–8] (Figure 1).

Figure 1. ECM alterations in aging heart tissue at different length scales.

Tissue organization changes with age due to cardiomyocyte hypertrophy, fibroblast activation, macrophage infiltration, and ECM remodeling (left panel). Zooming into the ECM (middle panel), aging results in structural and compositional changes (e.g., increased collagen V and VI, fibronectin), including at the fibril level (e.g., accumulation of advanced glycation end-products, AGEs, on collagen and glutathionylation of fibronectin fibers – right panel). Changes in ECM are accompanied by integrin activation and clustering that initiate mechanosignaling pathways. ECM: Extracellular matrix; Col: Collagen.

Aged-related dysfunctional states in cardiac cells include hypertrophy, senescence, reduced contractility (e.g., in cardiomyocytes), and myofibroblast activation (e.g., in fibroblasts, macrophages, and valvular interstitial cells). In general, ECM mechanical cues are transmitted through cells via activated integrins at the plasma membrane, leading to focal adhesion (FA) clustering, cytoskeletal tension, and the regulation of several downstream effectors including YAP/TAZ. In cardiomyocytes, these interactions occur at costameres – specialized adhesion structures that serve as primary sites for cell–matrix interactions. The costameric adhesome shares key FA proteins found in fibroblasts (e.g., talin and vinculin), which link cytoplasmic actin and intermediate filaments (e.g., desmin) to contractile sarcomeres [9]. Additionally, these adaptor proteins regulate actin cytoskeletal remodeling via the Rho GTPases/AKT/ERK pathways. In the presence of mechanical stress, FAK is activated and binds to Src family tyrosine kinases, disrupting RhoGTPase pathways. This causes cytoskeletal disorganization and activation of Ras/ERK/AKT pathways, ultimately leading to cardiac hypertrophy [10]. In matrix-regulating cardiac fibroblasts, the FAK/RhoA/ROCK pathway and Smad 2/3 promote myofibroblast activation and ECM remodeling [11]. Macrophages also function to indirectly remodel the matrix through integrin αvβ3 interaction with fibroblasts leading to Ca²⁺ influx via mechanosensitive ion channel Piezo1 [12]. This promotes myofibroblast differentiation of cardiac fibroblasts, independent of matrix mechanics. Figure 2 provides an overview of ECM-driven signaling pathways in cardiomyocytes, cardiac fibroblasts, and macrophages discussed throughout this review.

Figure 2. ECM-driven cellular signaling in cardiac aging.

Specific age-related ECM alterations activate distinct downstream signaling pathways in cardiomyocytes (top left panel), fibroblasts (bottom panel), and macrophages (top right panel) leading to tissue remodeling and fibrosis. In cardiomyocytes, young ECM maintains homeostasis through basal level integrin receptor activation and signaling, thereby preserving actin cytoskeletal dynamics and sarcomeric contracility. In contrast, aged ECM – characterized by increased stiffness stemming from accumulation of advanced glycation end-products (AGEs), excessive ECM deposition/abundance, and altered topography – activates the FAK/Src axis and Piezo1 pathways, resulting in impaired contractility, reduced Cx43, and hypertrophy. In macrophages, recruitment to fibrotic regions contributes to ECM remodeling. Glycoprotein non-metastatic melanoma protein B (GPNMB) is secreted, which binds to the ECM and transduces signals to cardiomyocytes and fibroblasts, inducing hypertrophy and ECM production, respectively. Macrophages also sense fragmented matrix through integrins, initiating cardiac repair processes. In fibroblasts, aged ECM promotes myofibroblast activation via several pathways: increased ECM stiffness triggers the integrin/FAK/Src axis, leading to the formation of highly contractile stress fibers; glutathionylated fibronectin induces integrin switching from α5β1 to αvβ3, activating the RhoA/PI3K/MAPK pathway; and altered ECM releases latent TGF-β, which can activate the canonical Smad2/3 pathway while Smad7 serves as a negative regulator. Topographical heterogeneity in aged ECM influences cytoskeletal remodeling and ECM gene expression through distinct signaling pathways, resulting in persistent myofibroblast activation or matrifibrocyte fibroblast states. ECM: Extracellular matrix; Itg: Integrin; PTM: Post-translational modification.

While much has been elucidated about cellular signaling cascades in the aging heart, contextual information is often lost, which is important because these processes are both region- and sex-specific. These two parameters are frequently not considered or understood in cell–matrix studies and biomaterials design. Here, we provide an updated window into how ECM alterations affect key cellular signaling pathways that lead to aberrant cellular responses and tissue remodeling in the aging heart. We propose the integration of new tools and the development of more specific models in ECM and cardiac aging research.

Distinct extracellular matrix cues drive cellular signaling in aging

Ligands

Excessive ECM deposition and remodeling in aging lead to dysfunction via reactive fibrosis. Significant matrisome alterations have been reported, including an increase in collagen I/VI, vitronectin, and fibronectin, while fibulin-5 is reduced [5,13] (Figure 1). The importance of ECM composition on cardiac cell function has been demonstrated in numerous in vitro studies. In Engineered Heart Tissues (EHTs) of human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs), matrix containing high levels of collagen types III and V vs. type I resulted in improved contraction-relaxation efficiency via several upregulated pathways including integrin/FAK/ERK/AKT/JNK/ROCK [5]. In myocardial infarction-conditioned chronologically aged hiPSC-CMs cultured with solubilized decellularized ECM derived from young human (<30 years) hearts, sustained beating frequency and reduced oxidative stress were found to be regulated by PI3K/AKT and NF-κB pathways, respectively [14]. Further, using hybrid hydrogel-decellularized ECM scaffolds with independently tunable stiffness and ECM, aged murine (18–24 months) cardiac fibroblasts were found to be more sensitive to ECM ligands vs. stiffness compared to young cells, which can be attributed to age-specific expression differences in mechanosensitive machinery [3]. Fibronectin increases in cardiac tissue with age, promoting FAK signaling and leading to activation and excessive ECM secretion in fibroblasts. While excess fibronectin is associated with fibrosis, cardiomyocyte cellular senescence has been linked to an unconventional pathway of fibronectin transcription and secretion through inhibited SENP1 deSUMOylation, which ultimately acts via P53 [15,16].

Ligand organization and availability have also been shown to dictate cell–matrix interactions in cardiomyocytes. Each ECM component has specific ligand arrangements, which can change due to alternative splicing, fibril crosslinking, fiber tensional state, and post-translational modifications (Figure 1). Using nanopatterned triangular substrates in which both local (on the triangles) and global (spacing between triangles) ligand densities can be tuned, the formation of stable adhesions in cardiomyocytes was found to depend on global RGD (fibronectin peptide) concentrations, while IKVAV (laminin peptide) required both local and global ligand densities [17]. This can be ascribed to RGD-integrin interactions causing talin stretching, thereby enhancing vinculin recruitment, adhesion complex formation, and strong adhesion forces.

Post-translational modifications of ECM, e.g., methylglyoxylation of collagen fibers by advanced glycation end-products (AGE) or site-specific glutathionylation of fibronectin fibers, cause alterations at the fibril level [18,19] (Figure 1). Such modifications can result in a higher fibril stiffness, increased lysine and arginine glycation, and altered surface charge of collagen fibrils which impacts integrin-ECM binding and ultimately mechanosensation [18] (Figure 2). On the other hand, glutathionylation of fibronectin alters fiber mechanical properties in fibrotic remodeling causing integrin switching from α5β1 to αvβ3 in activated fibroblasts [19] (Figure 2). Altered integrin binding can cause myofibroblast activation through a shift in downstream signaling. Lastly, age-related tissue remodeling can produce fragmented matrix molecules (matrikines) that attach to the ECM and elicit signaling events in matrix-modifying cells including fibroblasts and macrophages [20] (Figure 2). For instance, fragmented ECM can be sensed via integrin α5β1 in macrophages, leading to FAK/PI3K/AKT signaling that drives angiogenesis to aid in cardiac repair [21].

ECM stiffness

Tissue mechanics play an important role in regulating cardiac function, and it has been shown that passive tissue stiffness increases during aging due to both cell (e.g., titin) and ECM alterations [3,22]. At high strain levels, which exist in aged cardiac tissue, the ECM predominantly contributes to viscous forces [22]. This is important because an increase in ECM stiffness affects mechanosignaling through augmented integrin clustering and activation, vinculin enrichment, and F-actin re-organization [17] (Figure 2), which has been extensively reviewed elsewhere [23,24].

Enhanced tissue stiffness in a pressure overload model mimicking cardiac aging has been shown to activate mechanosensitive Piezo1 in murine (8–13 weeks) cardiomyocytes leading to Ca²⁺ influx [25]. This increased local Ca²⁺ concentration causes activation of the TRPM4 receptor, triggering the CaMKII-HDAC4-MEF2 pathway and ultimately resulting in cardiomyocyte hypertrophy (Figure 2). Mechanical stress can also activate EGFR pathways, causing cardiomyocytes to secrete paracrine signaling factors (e.g., cytokines and extracellular vesicles) that lead to cardiac fibroblast activation and ECM secretion [26]. While many phenotypic signatures overlap between pressure overload models and biological aging, an important difference is timeline, meaning cells are not allowed to adapt to the gradual increase in mechanics that occur throughout a lifespan, likely affecting epigenetic remodeling important in mechanical memory [27].

Cardiac fibroblasts are mechanosensitive, with higher stiffness causing myofibroblast activation. Using an on-demand stiffening 2D hydrogel, young rat (3 months) primary fibroblasts experiencing a stepwise increase in substrate mechanics underwent myofibroblast activation, cellular senescence, and acquired the senescence-associated secretory phenotype (SASP) [28]. In a stiffness-reversible 3D hydrogel system, neonatal rat (≤3 days) primary cardiac fibroblasts were exposed to sequential soft-to-stiff-to-soft states, and their activation status was found to be dynamic and directly correlated to mechanics, transitioning from quiescent-to-activated-to-quiescent. This dynamic phenotype was regulated by integrin-mediated mechanosignaling [29]. While this was observed in young cells, it important to note that cardiac fibroblast response depends on age state, with aged cells exhibiting reduced adhesion machinery and mechanosensitivity [3].

ECM topography

Cells can sense ECM topographical cues via FA complexes that regulate cytoskeletal arrangement and dictate cell alignment and shape. This is especially important in the heart where cardiomyocytes are organized in parallel bundles, which is required for efficient contractility. Multiscale topographical cues can be mimicked using nanolithography (e.g., micro-wrinkles and nanopillars in Polydimethylsiloxane, PDMS). In H9C2s and hiPSC-CMs, micro-wrinkles promoted alignment, nanopillars enhanced FA formation, and the combination of both topographies resulted in the highest maturation, including enhanced FA formation and higher lamin A/C and cardiac troponin T (cTnT) expression [30]. Fibrillar scaffolds (e.g., of silk fibroin) have been used to mimic heart architecture, revealing a fiber strain-specific response in rat cardiomyocytes [31]. At physiologically high strain, increased expression of TGF-β, collagen IA, vimentin, and α-SMA were found in hypertrophic cardiomyocytes, while higher strain levels found in pathological conditions resulted in disrupted sarcomeres, Z-lines, and cell–cell junctions [31].

Changes in topography have been shown to cause nuclear repositioning in fibroblasts [32], leading to activation of fibroblasts in the absence of TGF-β1³³. It was shown that 10 μm wide stripe micropatterns that are spaced 10 μm apart provide the optimal environment for fibroblast activation with higher cell elongation, FA formation, and cytoskeletal alignment. Combinatorial cues of topography and TGF-β1 lead to a more advanced activated state (i.e., cardiac matrifibrocytes) that regulate mature scar tissue through ECM remodeling [33,34] (Figure 2). Incorporating topography, elasticity, and ligands using an engineered cardiac mimetic matrix (CMM) platform found ECM property-specific regulation of cardiac ultrastructure [35].

ECM-immobilized proteins

Matrix-immobilized factors have been shown to regulate myofibroblast phenotypes and infiltration of cells into fibrotic tissue post-myocardial infarction. TGF-β1 is the most potent matrix-bound protein driving myofibroblast differentiation via the canonical Smad2/3 signaling pathway. To prevent excessive ECM deposition driven by myofibroblast activation, fibroblasts employ several regulatory mechanisms. Smad7 inhibits the Smad2/3 pathway, limiting collagen deposition and suppressing MMP2 activity, thereby preventing ECM degradation [36]. Alternatively, the ECM protein ADAMTSL3 reduces the bioavailability of TGF-β1 by interacting with fibrillin-1 and latent TGF-β1 binding protein-1 (LTBP1), effectively sequestering TGF-β1 and curbing its profibrotic effects [37] (Figure 2). Despite these regulatory attempts, fibrosis is further complicated in aging, with increased stiffness, altered ECM composition, dampened matrix remodeling, and impaired cell signaling. This dysfunctional matrix environment ultimately fuels a self-perpetuating fibrotic cycle.

TGF-β2 has been implicated in dilated cardiomyopathy through interactions with BAG3, which is essential for cardiac fibroblast proteostasis as it prevents excessive fibrogenic responses. In BAG3-deficient fibroblasts, the canonical SMAD2 pathway is upregulated, while SMAD3 and p38/MAPK pathways are suppressed [38]. Interestingly, while these fibroblasts lack α-SMA expression that is typical of myofibroblast activation [38], they are still capable of producing ECM, highlighting a subset of activated cardiac fibroblasts that contribute to ECM production independent of α-SMA expression.

Glycoprotein nonmetastatic melanoma protein B (GPNMB) has been found in fibrotic ECM and is secreted by macrophages [39,40] (Figure 2). Immobilized GPNMB adopts a different regulatory mechanism compared to mobile GPNMB by limiting ECM-specific gene expression in fibroblasts and sustaining contractility-related gene expression in cardiomyocytes, preventing cardiomyocyte hypertrophy via the AKT and ERK signaling pathways [40].

Getting more specific: sex and regional differences regulating ECM signaling

Sex differences

Mounting evidence suggest sex differences in the cardiac ECM and subsequent cell interactions with the ECM, which likely drive sex differences in heart function in both health and disease [41]. Sex differences in the heart are context-specific, and depend on an individual’s hormone status, sex chromosome composition, and age [42]. Cardiac sex differences in the ECM are apparent at the earliest stages of development and persist throughout aging.

The male sex is associated with increased left ventricle mass and higher extracellular volume, which predisposes them to increased fibrosis and elevated collagen abundance [43]. In pre-menopausal female hearts, increased estrogen signaling is attributed to suppressed gene expression of collagen protein and inflammatory factors [44]. After menopause, females lose the “protective” effect of estrogen in suppressing excessive matrix remodeling in the heart, causing increased accumulation of collagens after injury. In parallel with sex hormone activity in tissues, sex chromosome linked genes are important contributors to sex differences in ECM structures. In an ischemia/reperfusion injury mouse model, XX mice had higher vulnerability and infarct sizes to ischemia/reperfusion injury compared to XY mice, even after gonadectomy, suggesting cardiac scarring is a sex chromosome dependent process [45]. As such, sex hormones and sex chromosomes have independent and synergistic effects on cardiac tissue structures in health and disease [41].

On the cellular level, valvular interstitial cells (VICs) are the fibroblast-like cells that modulate progression of aortic valve stenosis (AVS) via fibrosis and calcification of the aortic valve ECM. Male and female VICs have sex-dependent gene and protein expression of myofibroblast markers (ACTA2, TGFβ1), ECM components (COL1A1, FN1, HAS3, CHSY-1), matrix metalloproteinases (MMPs), and tissue inhibitors of MMPs [46]. When VICs are cultured on hydrogel biomaterials that recapitulate the aortic valve matrix, genes coded on X and Y chromosomes have also been shown to modulate female-dependent myofibroblast activation and male-dependent osteoblast-like differentiation [47,48]. For instance, BMX and STS genes that escape X inactivation increase Rho/ROCK signaling activity to promote α-SMA expression [47], whereas the Y-linked UTY gene increased calcification associated genes including demethylases DEF6 and LLRN3, matrix remodelling and fibrocalcification associated genes CCBE1, DGKH, MBNL3, and ROBO1 [48]. VICs also have sex-dependent responses to inflammatory factors, showing elevated MAPK signaling activity driving female-specific myofibroblast activation [49,50]. Additionally, male and female VICs cultured on hydrogel biomaterials exhibit sex-dependent responses to low-dose drug combinations intended to inhibit the myofibroblast phenotype [51].

Sex-specific studies have also focused on cardiac fibroblast activation, with aged male (24–30 months) ECM promoting an injury-like matrix remodeling phenotype due to the enrichment of collagen IV, VI, and XII that is unique to murine male cardiac tissue [52]. This male-specific observation in ECM composition impairs the downstream kindlin/ERK/actin/α-SMA mechanosensing axis in cardiac fibroblasts, thereby diminishing mechanosensation required for myofibroblast maturation.

Regional differences

Cardiac tissue comprises distinct regions, including the ventricular walls, atrial walls, and septum, each with unique roles, tissue organization, and ECM composition. Recent studies have highlighted ECM-driven differences between the right ventricle (RV) and left ventricle (LV) under pressure overload and aging conditions [53]. Under pressure overload, the RV undergoes pronounced upregulation of glycoproteins including decorin, matrix-remodeling enzyme TIMP1, and collagen types I, III, and XIV, while the LV exhibits comparatively fewer changes [54]. Additionally, ECM-related gene expression is more heterogeneous in the RV, potentially reflecting localized adaptations to mechanical stress.

During aging, the RV and LV display distinct hypertrophic morphologies associated with different cellular pathways. The RV transcriptome reveals significant upregulation of metabolic and ECM-related pathways [53]. Notably, aging in the female RV involves more extensive ECM-driven signaling compared to males, underscoring the added complexity of sex-related differences in cardiac aging. Additionally, recent findings demonstrate that left ventricular small extracellular vesicles (LVVs) – which are cell-secreted nanoscale lipid vesicles containing DNA/RNA fragments, proteins, and metabolites that bind to the ECM – exhibit age-, sex-, and species-specific effects on fibrosis. While LVVs derived from females and young males (human <40 years old; corresponding mice age: 16 weeks) are anti-fibrotic, aged male (human >50 years old; corresponding mice age: 72 weeks) LVVs uniquely promote myofibroblast activation, driving cardiac fibrosis [55]. Overall, these findings highlight region- and sex-specific adaptations in cardiac tissue ECM during aging and stress.

Conclusions and future perspectives

Despite significant progress, existing in vitro models fail to fully capture the dynamic and multifaceted nature of ECM during aging. Here, we highlight the current state of knowledge for age-related transformations in cardiac ECM, emphasizing region- and sex-specific variations and their signaling consequences that underly cardiac aging dysfunction. Such insights hold promise for uncovering therapeutic targets to restore ECM homeostasis and improve cardiac function in aging populations.

There is still much to learn about how age-related cardiac dysfunction is specifically regulated through ECM alterations in native tissues. New applications of existing strategies and the development of emerging technologies will be required to uncover such knowledge [56,57]. Correlative technologies, including multiparameter spatial omics, are being harnessed to provide a more in-depth understanding of microenvironmental contexts and signaling regulation in tissues with high spatiotemporal resolution [58]. Further advancement in correlative tissue mapping will enable a direct linkage between mechanics, transcriptomics, proteomics, and metabolomics in native contexts [59].

Moreover, bioengineered tissues and organs have become increasingly important tools to understand key signaling events found in the heart [60]. The approach of culturing cardiac cells with engineered biomaterials in vitro has been used for decades yet historically have not considered specific factors including sex or tissue region as biological variables in experimental design [61]. New approaches in region- and sex-separating cardiac cells to investigate specific cellular phenotypes on biomaterials that also more accurately mimic the spatiotemporal aspects of aging cardiac matrix will shed insights into how unique biological variables contribute to cardiac responses [62]. This review underscores the urgent need for interdisciplinary approaches integrating biomaterials, mechanobiology, and bioengineering to tackle the complexities of ECM-driven signaling in the aging heart.

Data availability

No data was used for the research described in the article.