Drs. Santiago and Dangerfield contributed equally to the experimentation and analysis of data.
Cardiac fibroblast to myofibroblast differentiation in vivo and in vitro: Expression of focal adhesion components in neonatal and adult rat ventricular myofibroblasts
Article first published online: 9 APR 2010
Copyright © 2010 Wiley-Liss, Inc.
Volume 239, Issue 6, pages 1573–1584, June 2010
How to Cite
Santiago, J.-J., Dangerfield, A. L., Rattan, S. G., Bathe, K. L., Cunnington, R. H., Raizman, J. E., Bedosky, K. M., Freed, D. H., Kardami, E. and Dixon, I. M.C. (2010), Cardiac fibroblast to myofibroblast differentiation in vivo and in vitro: Expression of focal adhesion components in neonatal and adult rat ventricular myofibroblasts. Dev. Dyn., 239: 1573–1584. doi: 10.1002/dvdy.22280
- Issue published online: 17 MAY 2010
- Article first published online: 9 APR 2010
- Manuscript Accepted: 18 FEB 2010
- Heart and Stroke Foundation of Manitoba. Grant Number: G00309455
- Canadian Institutes of Health Research. Grant Number: G00300442
- cardiac fibroblast;
- extracellular matrix;
- matrix remodeling
In fibrosing hearts, myofibroblasts are associated with cardiac extracellular matrix remodeling. Expression of key genes in the transition of cardiac fibroblast to myofibroblast phenotype in post-myocardial infarction heart and in vitro has not been well addressed. Contractile, focal adhesion-associated, receptor proteins, fibroblast growth factor-2 (FGF-2) expression, and motility were compared to assess phenotype in adult and neonatal rat cardiac fibroblasts and myofibroblasts. Neonatal and adult fibroblasts undergo phenotypic transition to myofibroblastic cells, marked by increased α-smooth muscle actin (αSMA), smooth muscle myosin heavy chain (SMemb), extra domain-A (ED-A) fibronectin, paxillin, tensin, FGF-2, and TβRII receptor. Elevated ED-A fibronectin confirmed fibroblast to supermature myofibroblastic phenotype transition. Presence of myofibroblasts in vivo was noted in sections of healed infarct scar after myocardial infarction, and their expression is similar to that in culture. Thus, cultured neonatal and adult cardiac fibroblasts transition to myofibroblasts in vitro and share expression profiles of cardiac myofibroblasts in vivo. Reduced motility with in vitro passage reflects enhanced production of focal adhesions. Developmental Dynamics 239:1573–1584, 2010. © 2010 Wiley-Liss, Inc.
Stroma, composed of extracellular matrix (matrix) and mesenchymal cells, is a principal feature providing strong structural scaffolding in mammalian tissues. The principal stromal cell type are fibroblasts, but this designation belies their diversity and topographic differentiation from organ to organ including heart (Chang et al.,2002). For example, skin fibroblasts and vascular smooth muscle cells share the expression of desmin, and are thereby distinguished from virtually all other fibroblasts and myofibroblasts, including those in heart (Kalluri and Zeisberg,2006). Thus, the term “fibroblasts” designates a highly heterogenous group that exhibit distinct differentiated phenotypes in different tissues (Chang et al.,2002). The implications of these fundamental differences are unclear. Furthermore, the study of fibroblast and myofibroblast biology in specific organs is an important but relatively understudied area, especially in heart. Recent novel data indicates that ventricular fibroblast activation and cardiac fibrosis are primary events in ventricular remodeling, rather than a secondary response to cardiomyocyte injury (Thum et al.,2008). Thus, the traditional role of cardiac fibrosis as a secondary disease modifier has recently, and for the first time, been called into question (Thum et al.,2008), and the need to establish the specific behavior and expression patterns of genes that characterize cardiac myofibroblasts is apparent.
Matrix components of the healthy heart are produced by interstitial cardiac fibroblasts (Eghbali et al.,1988). These cells maintain a relatively slow turnover of fibrillar collagens (Weber,1997) in normal conditions, but may respond to both mechanical loading (Wang et al.,2003) or transforming growth factor-β1 (TGF-β1) stimulation (Desmouliere et al.,1993; Petrov et al.,2002) by a switch to a myofibroblastic phenotype wherein they express α-smooth muscle actin (αSMA; Leslie et al.,1991), synonymous with increased contractile force (Hinz et al.,2001). αSMA expression is increased in myofibroblasts in fibrotic hearts subjected to pressure or volume overload or in the infarct scar of post-myocardial infarction (post-MI) hearts (Weber,1997). Causal factors in this conversion are compressibility of the substrate when ventricular fibroblasts are plated in vitro (Masur et al.,1996) and overexpression of R-Smads (Bujak et al.,2007). Enhanced contractility that attends this protein's expression is believed to be important in allowing these cells to contract while bound to matrix collagens and other proteins, thereby allowing for physical remodeling of the matrix itself (Arora and McCulloch,1994). Thus, myofibroblasts are the primary mediators of wound healing in the damaged ventricle, and we have previously demonstrated that they are the dominant cell type in the infarct scar (Peterson et al.,1999). Investigation of these cells in hypertrophied hearts is clinically relevant as they contribute to wound healing, matrix remodeling, and eventual cardiac fibrosis through the elevated production of fibrillar and nonfibrillar collagens (Cleutjens et al.,1995; Shamhart and Meszaros,2010).
Myofibroblasts exhibit a stellate form with nuclei marked by nucleoli, and feature prominent stress fibers with highly developed focal adhesions (Tomasek et al.,2002; Kalluri and Zeisberg,2006). These cell-to-matrix junctions, consisting of myofilament and fibronectin filament systems that converge on a discrete cell-surface plaque, which in turn are made of paxillin, vinculin, and tensin, among other proteins. Unlike other cells, myofibroblasts lack a prominent basal lamina and thus focal adhesions (and their component proteins) are useful markers for discriminating these from other cells (Eyden,2001b). Myofibroblasts also express prominent αSMA myofilaments and associated vimentin proteins that form stress fibers within the cytoskeleton (Kalluri and Zeisberg,2006). Although we and others have reported the primary cardiac ventricular fibroblasts' conversion to myofibroblastic phenotype in vitro, this phenomenon remains a controversial topic in cardiac biology, and furthermore, the characterization of focal adhesion component expression is not well characterized in these cells.
Focal adhesion-associated proteins participate in cellular events that confer myofibroblast contractility and migration. Myofibroblasts may be characterized by increased expression of embryonic smooth muscle myosin heavy chain (SMemb) (Frangogiannis et al.,2000) and extra domain-A (ED-A) fibronectin (Borenstein et al.,2003); the latter protein is bound to fibronexus adhesion complexes in vivo and supermature focal adhesions in vitro (Tomasek et al.,2005). Thus, ED-A fibronectin is critical for linking the extracellular matrix to intracellular cytoskeleton, and for exerting mechanical traction on the matrix by means of cellular contraction. Myofibroblasts migrate to the infarct zone, restoring cellularity (Norman,2004). Their contraction confers matrix remodelling by imparting tensile force to the matrix, opposes retractile force, promotes scar contraction, activates latent TGF-β1, and reorients collagen fibrils (Grinnell,1994; Lijnen et al.,2003; Wang et al.,2003; Arany et al.,2006; Wipff et al.,2007). Although αSMA is a useful marker for the fibroblast to myofibroblast switch (Tomasek et al.,2002) in noncardiac tissues, we hypothesized that cardiac myofibroblasts are characterized by up-regulation of focal adhesion-associated proteins (including the ED-A splice variant of fibronectin), specifically in myocardial infarct scar and in vitro. We also wished to assess the expression of growth factors implicated in fibroblast phenotype regulation. These include fibroblast growth factor-2 (FGF-2), a heparin-binding protein synthesized as high molecular weight (21.5–22 kDa) and low molecular weight (18 kDa) isoforms, termed hi- and lo-FGF-2, respectively; and TGF-βRI and TGF-βRII (Tseng et al.,1999; Greenberg et al.,2006). Furthermore, we sought to compare the phenotype of adult cardiac fibroblasts and myofibroblasts with that of neonatal fibroblasts and myofibroblasts, and compare those cultured cells with those in cardiac infarct scar. The current data indicate that primary adult and neonatal ventricular passage 0 (P0) fibroblasts undergo rapid and consistent transition to the myofibroblastic phenotype upon plating in our culture conditions. The data also extend our previous in vivo studies of cardiac myofibroblast phenotype in vivo. Finally, we show that myofibroblasts of the myocardial infarct scar are closely matched phenotypically to those cultured in vitro, and that in heart, ED-A fibronectin, a focal adhesion-associated protein, is highly expressed in both experimental systems. We suggest that up-regulated focal adhesion-associated proteins provide reliable markers to identify cardiac myofibroblasts.
Fibroblast to Myofibroblast Conversion in Culture on Rigid Matrix and Purity of Cultures
We addressed the consistency of cardiac fibroblast phenotype shifting in vitro, using cell culture on standard plastic plates (Masur et al.,1996; Wang et al.,2003; Freed et al.,2005), and have compared this trend in adult and neonatal cells. Our findings summarize the expression of key marker proteins which are up-regulated in the myofibroblastic phenotype. Figure 1B shows the expression of vimentin, αSMA, and SMemb in adult ventricular cells using Western blot analysis. The expression of these three marker proteins were up-regulated either in the first or second passage (P1 or P2) in cultured adult ventricular cells and then remained increased in P3 when compared with expression in quiescent P0 fibroblasts. Figure 1C provides representative immunofluorescence data to show the expression of vimentin and lack of desmin expression, which indicates purity of fibroblast culture and increased αSMA expression (fibroblast to myofibroblast shift). Neonatal cells (Fig. 1A) showed similar directionality in expression patterns of the three proteins assayed, but with a delayed increase of vimentin and αSMA, e.g., vimentin increased only in P3 cultures and αSMA in P2 and P3 vs. P0 controls. SMemb was increased eight-fold vs. P0 control in adult myofibroblasts and four-fold compared with P0 control in neonatal myofibroblasts. Together, these results indicate that neonatal cells may exhibit a slight delay in their conversion to myofibroblasts compared with adult cells. Alternatively, development is still proceeding in the neonatal cardiac cells, and are less mature with different responses from adult cells; however, this possibility requires further investigation.
Upregulation of Focal Adhesion Proteins in Myofibroblasts
Morphologically, myofibroblasts are characterized by the presence of stress fibers or actin bundles that terminate at the plasma membrane-bound focal adhesions, thus we examined focal adhesion proteins as a function of passage number. Fibronectin protein exists as several splice variants and the ED-A variant is of particular interest in distinguishing myofibroblastic phenotype. We have observed a ∼262-kDa band in human cardiac myofibroblasts (the expected molecular mass), however, using the same antibody in rat cardiac myofibroblasts we found major bands at 115, ∼82, and ∼52 kDa (data not shown). Thus, rat ED-A fibronectin appears as three specific bands of comparable intensity, likely as proteolytic products. These observations were then applied to the current dataset and the ∼52 kDa band used to compare the focal adhesion protein expression of cells from different passage numbers. Figure 2 shows the expression of paxillin, vinculin, tensin, and ED-A fibronectin in Western blot analyses. Adult ventricular cells (Fig. 2C,D) show a significant increase of paxillin, tensin and ED-A fibronectin in second and third passages when compared with expression in P0 fibroblasts. Vinculin expression was unchanged among passages in adult cells however we found that in neonatal ventricular cells (Fig. 2A,B), vinculin was significantly increased in all passages vs. P0 fibroblasts. Paxillin, tensin, and ED-A fibronectin all show similar increases in expression with later passages (P2 and P3). Neonatal myofibroblasts showed an eight-fold increase compared with P0 fibroblasts in terms of vinculin, tensin, and ED-A fibronectin, while adult myofibroblasts showed a two- to four-fold increase vs. P0 cells. Paxillin expression was similarly increased in both adult and neonatal myofibroblasts.
Expression of Growth Factors and Receptors in Fibroblasts and Myofibroblasts
It is well established that the treatment of fibroblasts with TGF-β1 induce myofibroblast differentiation by activating TGF-β1 receptors; however, to our knowledge, there is no known report showing the endogenous expression of these receptors in myofibroblast differentiation in vitro in the absence of stimulation with this factor. We examined the expression pattern of TGFβ RI and RII in both neonatal and adult myofibroblasts as a function of passage number (Fig. 3A,C). In neonatal myofibroblasts (Fig. 3A), no significant changes were determined for TGFβ RI; however, there was an increase in TGFβ RII expression as early as P1 vs. P0 cells and the upregulation of TGFβ RII is apparent in all passages (P2–P3). This trend is similar in passaged adult myofibroblasts compared with P0 cells. We also examined expression of the 18-kDa low molecular weight lo-FGF-2 and the 21- to 22-kDa high molecular weight hi-FGF-2 isoforms in myofibroblast differentiation as a function of passage number. Increased relative expression of both hi- and lo-FGF-2 in P1–P3 compared with P0 samples was observed in both neonatal (Fig. 3B) and adult myofibroblasts (Fig. 3D) vs. P0 fibroblasts.
Increase in Collagen Synthesis Is Correlated With Fibroblast to Myofibroblast Transition
We examined the functionality of primary myofibroblasts to produce extracellular matrix-bound collagen and compared this with quiescent fibroblast collagen synthesis. First, we examined the levels of intracellular procollagen in adult cells and found its expression was significantly increased with passage, reaching a maximal level at P2 (data not shown). This increase is reflected using immunofluoresence staining of procollagen (Fig. 4A) at later passages. As well, secreted mature type I collagen was increased as early as P1 and maintained until P3 (Fig. 4B) vs. P0 samples in both neonatal and adult cells indicating that both neonatal and adult P1 to P3 cells are hypersynthetic for type I collagen vs. P0.
Fibroblast and Myofibroblast Migration Assay
To examine whether passage in culture affected the motility of cells, we carried out Transwell experiments using two different concentrations of cardiotrophin-1 (CT-1), a known chemokine. Cell count was taken as a measure of cellular motility. Figure 5 illustrates the concentration-dependent effects of CT-1-induced-motility. Using adult ventricular P0 fibroblasts as well as adult P1 and P2 myofibroblasts (2 × 105 cells/insert) starved in serum-free media, we found that with no treatment of cells, no significant difference among the three groups was observed. Under these conditions, we ruled out the effects of cell proliferation in final cell counts. Conversely, with 1 and 10 ng/ml CT-1 in the lower well, we found a significant increase in P0 fibroblast migration vs. P1 or P2 myofibroblasts (*P < 0.001). Furthermore, we found a significant increase in the motility of cells in P1 vs. P2 myofibroblasts (#P < 0.05).
Expression of Myofibroblast Markers and ED-A Fibronectin in Myocardial Infarct Scar
To demonstrate that cardiac myofibroblasts cultured in vitro are phenotypically similar to those found in healed cardiac infarct scar in vivo, immunofluorescence studies were carried out to characterize those cells in healed myocardial infarct scar (4 weeks after myocardial infarction). Myofibroblasts stained for key protein markers, e.g., αSMA, discoidin domain receptor 2 (DDR2), SMemb, and counterstained with Hoechst to visualize nuclei are shown in (Fig. 6). Furthermore, we examined the expression of ED-A splice variant of fibronectin in frozen tissue sections of infarcted rat hearts, as this form of fibronectin directly binds to focal adhesions and their associated proteins (Tomasek et al.,2002), allowing for a direct connection between the contractile myofibroblast and matrix in vivo. Skin fibroblasts and vascular smooth muscle cells express desmin, whereas cardiac myofibroblasts do not (Kalluri and Zeisberg,2006), and thus a lack of desmin expression in heart tissues is associated with fibroblastic cell types. We found that desmin-negative cells of the infarct scar expressed both αSMA and DDR2, in focal regions of the infarct scar (Fig. 6A–A3). Furthermore, these cells also stained positive for SMemb (Fig. 6E,E1) and ED-A fibronectin (Fig. 6F,F1) in the absence of desmin expression.
Appearance of Collagen in the Infarct Scar
Finally, in a histological study using “breadloaf” thin slices of myocardium stained with Masson's trichrome, we provide data to confirm the presence of extensive collagen deposition in the 4-week post-MI heart (Fig. 7A,B; sections of 4-week noninfarcted control myocardium, and 4 week post-MI infarct scar, respectively). This study allowed us to distinguish parenchymal muscle cells from surrounding connective tissues. In these sections, pink muscle fibers and cytoplasm are resolved from blue-stained extracellular matrix (mainly fibrillar collagens). A relatively small amount of blue-stained collagen was found in healthy control heart sections (Fig. 7A). Extensive collagen deposition is seen in the infarcted heart section (Fig. 7B). The regions of elevated collagen deposition positively correlate to the localization of cardiac myofibroblasts.
The architecture and functional design of the cardiac stroma, composed of extracellular matrix and mesenchymal cells, provides a structural scaffold for linkage of cardiac myocytes, and serves to orient myocytes in three dimensions for nominal cardiac pump function (Ott et al.,2008). Fibroblasts exhibit marked phenotypic diversity and display distinct and characteristic transcriptional patterns, suggesting that fibroblasts residing in distinct organs should be considered distinct differentiated cell types (Chang et al.,2002). In this respect, we observed that both in vitro and in vivo cardiac myofibroblasts are of a similar phenotype, and that this phenotype is likely unique to heart (data not shown). In diseased heart, cardiac fibroblasts become activated not only in parallel with myocyte damage but also as a primary mechanism of disease; however, this paradigm has only very recently been recognized (Thum et al.,2008). To facilitate the identification of cardiac myofibroblasts, we compared the expression of established and novel phenotypic markers in these cells. Myofibroblasts resemble fibroblast cells by having an extensive rough endoplasmic reticulum and Golgi apparatus, but in contrast to fibroblasts, myofibroblasts display de novo expression of αSMA, supermature focal adhesions, and SMemb (Frangogiannis et al.,2000; Tomasek et al.,2002). Our data indicated that adult and neonatal cardiac fibroblasts acquire contractile and synthetic features during differentiation to myofibroblasts, but that this phenotype acquisition was somewhat delayed in the neonatal vs. adult cells, despite the similar directionality of change. We suggest that the consistent elevation of paxillin and tensin by P3 in neonatal cells and P2 in adult cells is indicative of the upregulation of focal adhesion component proteins and along with ED-A fibronectin (known to bind directly to focal adhesions and to serve as connectors to matrix), focal adhesion maturity (Kalluri and Zeisberg,2006). Notably, these specific proteins (including ED-A) are also present in myofibroblasts of the healed infarct in vivo, as shown in our current data. Furthermore, healed infarcts in 4-week post-MI cardiac sections were characterized by the presence of extensive collagen deposition compared with controls. These events paralleled the onset of appearance of the myofibroblast markers αSMA and SMemb (as well as vimentin). A striking reduction in cellular motility was also observed with increased passage and this correlates closely with elevated focal adhesion component protein expression in vitro. We suggest that in cardiac fibroblasts and myofibroblasts, increased focal adhesions lead to a relatively immobile cell that retains its hypersecretory function. Elevated αSMA contributes to the characteristic stellate cell shape and form stress fibers, as the cells contract on substrate. Thus, while both adult and neonatal fibroblasts consistently undergo phenotype switching to myofibroblasts, these processes show subtle differences in each group. This notwithstanding we suggest that neonatal myofibroblasts may be a useful model of the adult cells in vitro.
We noted a marked increase in SMemb in both neonatal and adult cardiac ventricular cells with passage. Although it seems logical to suggest that SMemb is a more robust marker for ventricular myofibroblast switching compared with other target proteins examined, the precise function of this nonmuscle myosin in these cells is relatively undefined. Conversely, the focal adhesion—associated proteins such as paxillin and tensin—are major scaffolding and integrin-linked protein components of focal adhesions, respectively (Geiger et al.,2001) and are thus a direct marker for the presence of focal adhesions. Due to the relatively weak staining of myofibroblasts for basal lamina proteins such as type IV collagen and laminin as well as the presence of prominent focal complexes and focal adhesions in these cells (Eyden,2001a), and the current data, we suggest that the expression of their components are useful tools for reliable identification of these cardiac cells. Although we observed some modulation of the elevation of tensin expression in neonatal cardiac myofibroblasts, we expect that this trend reflects a damping of expression with time. As expression of markers other than tensin remain significantly elevated through P3, these cells are not likely to revert back to a noncontractile phenotype. Nonetheless, as myofibroblast phenotype reversion is emerging as a feasible topic for investigation (Greenberg et al.,2006) and represents a putative therapeutic strategy, it requires further investigation.
Pathways That May Influence Cellular Differentiation
The 18-kDa lo-FGF-2 isoform, has been implicated in the reversal of myofibroblast phenotype, as per the work of Greenberg and colleagues (2006). There is as yet no information on the activity of the hi-FGF-2 isoform. In addition, there is no information as to the relative expression of the FGF-2 isoforms in cardiac fibroblasts and myofibroblasts. In both neonatal and adult cells, we noted significant upregulation of both types of FGF-2 isoforms during the switch to the myofibroblastic phenotype. We suggest that this increase allows these cells to potentiate the signal for phenotype modulation. Furthermore, hi- and lo- MW FGF-2 variants may lend pleitropism (Okada-Ban et al.,2000), with each isoform exerting markedly different or opposing function(s), and these effects may serve to balance one another. Our data indicate increased FGF-2 expression observed in successive passages of myofibroblasts—increased passage was also accompanied with increased αSMA expression. This finding is in apparent disagreement with reports by several investigators including Greenberg et al. (2006) using mouse embryonic fibroblasts, or Ishiguro et al. (2009), who used skin-derived fibroblast cell lines and have found that FGF-2 reverses or prevents conversion to myofibroblast phenotype, as determined by reduced αSMA expression. In cardiac myofibroblasts, we did not observe this relationship, but rather found that αSMA expression is unresponsive to FGF-2 treatment (data not shown). This is very likely a reflection of differences between fibroblasts derived from different tissues and/or different developmental stages (Chang et al.,2002; Kalluri and Zeisberg,2006). For example, skin-derived fibroblasts are reported to express desmin, which is not expressed by fibroblasts from other organs (Kalluri and Zeisberg,2006).
Other subcellular signaling systems may subserve to regulate fibroblast phenotype. In this regard, TGFβ1 is well established as a player in the transition of fibroblasts to myofibroblasts, and Petrov and colleagues suggest that this, rather than direct stimulation of collagen, is the major cause for increased collagen secretion by these cells in culture (Petrov et al.,2002). Our findings indicate that TGFβ RII is significantly increased with passage, and we suggest that as the actual binding of TGFβ1 may be the rate limiting step, that this change is sufficient for potentiation of the TGFβ1-Smad signal. We have previously noted increased phosphorylation of R-Smad2 in P1 cells within 5 min of TGFβ1 treatment, and suggest that the transition of fibroblast to myofibroblast provides an ongoing stimulus to maintain the latter phenotype in culture (Wang et al.,2007). In this respect, neonatal and adult fibroblasts appear to express similar subcellular Smad-dependent signaling pathways (Colwell et al.,2006).
Functional Implications of Upregulated Focal Complex/Focal Adhesion Proteins
Upregulation of focal adhesion proteins may provide the basis of inherently reduced motility, as cells become anchored to the surrounding extracellular matrix. Using the Transwell system to address cell motility, we have observed that freshly isolated P0 adult cardiac ventricular fibroblasts are significantly more motile than P1 myofibroblasts and that P2 cells are less motile than their P1 counterparts, using cardiotrophin-1 as the stimulus (Fig. 5). In this regard, paxillin is a key focal adhesion protein, as it is sandwiched between integrins and the extracellular matrix, and intracellularly, the actin cytoskeleton of cells (Turner,2000). The increase of paxillin was paralleled by increased tensin in both neonatal and adult cells however the lack of significant change in vinculin expression in adult myofibroblasts was unlike the trend in its expression in neonatal myofibroblasts. This is one of the few differences in the comparison between the adult and neonatal myofibroblasts in vitro. The current data also support the suggestion that cardiac myofibroblasts are hypersynthetic for collagens and cytokine signaling pathway components. Finally, ED-A fibronectin is increased in all cultured cardiac myofibroblasts as well as strongly expressed in vivo in myofibroblasts from healed infarct scar, and is likely involved in tethering these cells to surrounding matrix by means of direct connections to focal adhesions at the cell surface.
In summary, we sought to identify phenotype transition changes specific to cardiac ventricular fibroblasts, as it is likely that these cells are unique when compared with fibroblasts from other organs. Data from the current study indicate that both neonatal and adult cultured cardiac ventricular fibroblasts differentiate into myofibroblasts when seeded at low density. During this transition, fibroblasts undergo phenotypic changes leading to expression of contractile proteins such as αSMA and SMemb as well as ED-A fibronectin in vitro. We have shown robust ED-A fibronectin expression in vivo and in vitro. ED-A fibronectin has been shown to interact with focal adhesion proteins and to localize within the focal adhesion complexes (Tomasek et al.,2002), Thus, the increase in ED-A fibronectin within the scar as shown in Figure 6F,F1, together with the in vitro data showing that ED-A fibronectin increases parallel those of vincullin, paxillin, and tensin. This would indicate an increase in focal adhesion complexes in vivo. Increased abundance of focal adhesions is correlated to decreased cardiac myofibroblast motility. Both neonatal and adult cardiac myofibroblasts share expression of key marker proteins, as seen in vivo. Thus, cardiac fibroblasts are a versatile cell population which may reproducibly converted to myofibroblasts in vitro under our conditions, and will be useful models for further study.
Isolation of Adult Cardiac P0 Fibroblasts
Ventricular fibroblasts were isolated from the hearts of 150–200 g adult male Sprague-Dawley rats as previously described (Hao et al.,2000). Briefly, hearts were subject to Langendorff perfusion at 37°C with medium containing collagenase (Worthington Biochemical Corporation, Lakewood, NJ) for 20–25 min. Collagenase was neutralized by addition of an equal volume of medium containing 10% fetal bovine serum (FBS) and liberated cells were collected by centrifugation. Cells were resuspended in fresh medium containing 10% FBS and plated on 75 cm2 culture flasks at 37°C with 5% CO2 for 3 hr. Nonadherent cells (myocytes) were removed by changing the culture media and adherent cells (mainly fibroblasts) were incubated in medium containing 10% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin, and 100 μM ascorbate. These cells were labeled as P0 and passaged three times accordingly (P1–P3) for experiments. The purity of these cells was 95%, using routine phenotyping methods as previously described (Peterson et al.,1999).
Isolation of Neonatal Cardiac Fibroblasts
Primary cultures of neonatal cardiac fibroblasts were isolated from 1-day-old Sprague-Dawley rat pups (36 pups/preparation) according to standard protocols (Doble et al., 1996, 2000). Briefly, hearts were placed in a Petri dish with cold phosphate-buffered saline (PBS) containing 3.5 g/L of glucose. The hearts were minced to approximately 1 mm3 or small enough to pass through a 10-ml pipette tip. The heart tissue was digested at 37°C in collagenase. The digestion was repeated six times, and the cells were pooled in a bottle with 10 ml of FBS to inactivate the enzymes. After the sixth digestion, the cells were filtered through a Nytex membrane using a syringe filter; and centrifuged at 2,000 rpm for 5 min. The pellet was resuspended in 1× ADS buffer and once again filtered through a Nytex membrane. The cell suspension was layered onto a discontinuous Percoll gradient (Amersham Biosciences) and centrifuged at 3,500 rpm for 30 min. The upper band of cells was removed and plated in 100-mm cell culture dish and labeled as P0 cells. The cells were passaged one (P1), two (P2), or three (P3) times.
Experimental Rat Model of MI
MI model was produced in male Sprague-Dawley rats (150–175 g body mass) by surgical occlusion of the left coronary artery, as described previously by us (Dixon et al.,1990). The mortality of the animals operated on in this manner was <10% within 48 hr. Experimental animals used were sacrificed at 2 weeks and 4 weeks. Cardiac scar tissues and tissue bordering the infarct scar were isolated from left ventricle regions and frozen in embedding medium for sectioning. Non infarcted tissue was also collected from these hearts.
Total Cell Lysates Protein Extraction and Protein Assay
Total cell lysates were obtained by scraping cells directly into 1× sodium dodecyl sulfate (SDS) buffer (1% SDS, 50 mM Tris-HCl pH 6.8, and 10% glycerol) with 1:100 dilutions of protease (PIC) and phosphatase (PPIC I, and PPIC II) inhibitor cocktails. Lysates were boiled for 5 min and sonicated briefly, setting the frequency rate at 40 Hz. The samples were then centrifuged at 14,000 rpm for 5 min at 4°C with a micro-centrifuge. The insoluble pellets were discarded and the supernatants were transferred into autoclaved Eppendorf microcentrifuge tubes. Bicinchoninic Acid (50 ml) with copper sulfate (1 ml) was used to determine the protein concentration of the supernatants (BCA technique; Smith et al.,1985). Different known concentrations of bovine serum albumin (BSA) were used to obtain a standard curve. The samples were measured in triplicates and the absorbance was read using a spectrophotometer at a wavelength of 562 nm.
Western Blot Analysis of Target Proteins
For the total cell lysates, the samples were mixed in a 1:5 ratio (v/v) with 5× SDS loading buffer (10% SDS, 50% glycerol, 0.5 M dithiothreitol, 300 mM Tris-HCl pH 6.8, 0.005% bromo phenol blue). The samples (10–50 μg of proteins) were boiled for 5 min before loading onto a 7.5–15% SDS-polyacrylamide gel electrophoresis gel. Broad-range molecular weight standards and pre-stained molecular weight markers were also included in the gel. Separated proteins were electrophoretically transferred onto a 0.45 μM polyvinylidene difluoride membrane for 1 hr at 300 mA for small proteins or for 2 hr at 500 mA for proteins greater than 100 kDa using a buffer containing 20% methanol, 192 mM glycine, and 25 mM Tris base. Membranes were stored in a Tris-buffered saline (10 mM Tris-HCl pH 7.6 or 8.0 and 150 mM NaCl) with 0.1% Tween-20 (TBS-T). Before probing, blots were checked for equal protein loading between lanes using Ponceau S stain. Also, the molecular weight standards were marked while the blots were stained with Ponceau S. The membranes were blocked in a TBS-T solution containing 5–10% dried nonfat (skim) milk powder for a period of 1 hr at room temperature on an orbital shaker. The membranes were briefly rinsed with two changes of TBS-T. The blots were then incubated with primary antibody in 1% nonfat skim milk powder in TBS-T overnight at 4°C with constant agitation or for 1 hr at room temperature. Membranes were briefly washed with two changes of 1% milk in TBS-T. Then the membranes were washed again for 15 min once, followed by 3 washes for 5 min at room temperature on an orbital shaker. Secondary antibody consisted of anti-mouse and anti-rabbit immunoglobulin conjugated to horseradish peroxidase (HRP) was then applied at 1:10,000 dilution in TBS-T containing 1% skim milk and incubated for 1 hr at room temperature on an orbital shaker. Following the secondary antibody, the membranes were washed again following the same methods as described for the primary antibodies except the membranes were washed in TBS-T only. Protein bands on Western blots were visualized by ECL Plus (Amersham) according to the manufacturer's instructions, and developed on film. Exposures of film (Kodak) ranged from only a few seconds to over 20 min depending on the samples. The bands were normalized against glyceraldehyde-3-phosphate-dehydrogenase (GAPDH) or Actin signal; and used to confirm even protein loading. Band intensity was quantified using a CCD camera imaging densitometer (GS670, Bio-Rad Laboratories [Canada] Ltd. Mississauga, Ontario).
Cells between P0 and P2 were seeded onto coverslips in six-well dishes and allowed to attach overnight in media containing 10% serum. The cells were rendered quiescent in serum-free media for 24 hr. Media was removed; cells were rinsed with 1× PBS and fixed with 4% paraformaldehyde. Cells were permeabilized with 0.1% Triton X-100 and incubated with primary antibodies (probed for pro-collagen type I, αSMA, vimentin), biotinylated secondary antibodies, and streptavidin fluorescein isothiocyanate (FITC). Nuclei were stained with Hoechst 33342. The cells were visualized with epifluorescent microscopy with appropriate filters (Nikon Canada).
Tissue was prepared according to procedures as described (Yusuf et al.,2004). Briefly, hearts, excised from animals killed at 4 weeks after left ventricular coronary ligation surgery were used to cut (Microm HM 550 cryotome) 7-μM-thick transverse sections across the isolated left ventricular scar area. Sections were incubated with primary antibodies diluted to 5–10 μg/ml, as per manufacturer's instructions, in 0.5% BSA in PBS, and incubated overnight in the cold. To visualize antigen-antibody complexes, sections were incubated for 1 hr at room temperature with secondary antibodies fluorescein-labeled anti-mouse; and Texas-Red-labeled anti-goat antibodies, (Jackson Laboratories, Bar Harbor, ME), diluted as per manufacturer's instructions. After washing, sections were mounted using Prolong Gold mounting medium (Invitrogen, Carlsbad, CA). Sections were observed using a Zeiss Axiovert 200M epifluorescence microscope with mechanized stage and Z-axis image acquisition capability. The Axiovision program, Axiocam camera were used for image acquisition.
Tissue was prepared according to procedures as described (Yusuf et al.,2004) with the exception of tissue fixation in 95% ethanol. Briefly, hearts were excised from animals killed at 4 weeks after left ventricular coronary ligation surgery. Isolated left ventricular tissue was frozen in embedding medium at −80°C to be used for sectioning. Frozen tissue block was cut in 7-μM-thick transverse sections across the ventricular scar area using a Microm HM 550 cryotome. Masson's trichrome staining was carried out by members of the Department of Pathology at the St. Boniface General Hospital (Winnipeg, Canada).
Measurement of Procollagen Type I N-terminal Peptide (P1NP) Secretion Using Enzyme Immunoassay
The P1NP enzyme immunoassay (EIA) was carried out using the rat/mouse P1NP EIA kit from ImmunoDiagnostic Systems (IDS), allowing for the accurate measure of secreted collagen type I to cell culture media. Collagen synthesis in myofibroblasts is subject to turnover of monomeric protein components and numerous post-translational modifications, and thus secretion of the mature protein is a reliable indicator of net collagen synthesis. P1, P2, and P3 cardiac myofibroblasts were grown in 100-mm cell culture dishes until they reached 70% confluency. Cells and media were scraped from the culture plates and 5 μl of homogenized sample (cells and media) was added to the 96-well antibody coated plate in duplicate and 45 μl of sample diluent was added to each sample. 50 μl of P1NP Biotin was added to all standards and samples shaken at 500–700 rpm for 1 hr at room temperature. All wells were then washed three times with 250 μl of wash solution and 150 μl of enzyme conjugate was added to all of the wells and allowed to sit at room temperature for 30 min. The wells were then washed again three times with wash solution and 150 μl of TMB substrate is added to all wells and allowed to incubate 30 min at room temperature. Finally, 50 μl of stop solution was added and the plate read on a spectrophotometer at 450 nm within 30 min. The OD readings were used to calculate collagen/sample using the standard curve.
Cellular Migration Assay: The Transwell System
The Costar Transwell assay consists of two chambers separated by a filter through which cells migrate. Cells respond to a chemotactic gradient by moving in three dimensions and squeeze through the pores of the specific filter (8 μm in this study). Our previous experience with cardiotrophin-1 led us to use this chemokine as a stimulus for our cells (Freed et al.,2003,2005). Nonetheless, our results confirm the efficacy of our experimental conditions. Chemotaxis of adult rat cardiac fibroblasts (P0) and myofibroblasts (P1 or P2) were determined as previously described with minor alterations (Jo et al.,2000). In lower chambers of 6-well Transwell plates, chemokines were diluted to specified concentrations in 2.5 ml serum-free DMEM/F12. Cells were plated into inserts, which were superimposed to (and separated from) the lower chamber by a polycarbonate membrane containing 8μm pores. For experiments using P1 and P2 myofibroblasts cells were passaged, counted with a hemocytometer, and 2 × 105 cells/well were plated directly into inserts. For experiments using P0 fibroblasts the pellet, obtained after cells were freshly liberated from rat hearts, was resuspended in serum-free media then 2 ml was added into each insert. After 2 hr, inserts were gently washed twice with PBS to rid nonadherent cells followed by addition of 1.5 ml of serum-free media. Transwell plates were incubated at 37°C for 24 hr. Cells which had migrated through the membrane toward the chemoattractant gradient became adherent to the underlying membrane and at the bottom of the lower well. Media were carefully aspirated from both the inserts and wells and washed once with 1.5 ml of PBS followed by addition of 1.5 ml of trypsin to the lower well. Plates were incubated for 5 min, gently agitated to detach adherent cells, and then neutralized with equal volumes of DMEM/F12. Inserts were removed and the cell suspension was diluted in filtered PBS, and counted with a Model ZM Coulter cell counter (Beckman Coulter, Fullerton, CA). The number of counted cells is representative of rate of migration. Each experimental group was performed in duplicate.
The antibodies used were monoclonal anti-FGF-2, monoclonal anti-vimentin, DDR2 (discoidin-domain receptor 2), anti-TGFβ-R1, anti-TGFβ-RII, and monoclonal mouse anti-GAPDH (Santa Cruz Biotech., Santa Cruz, CA); monoclonal anti-α-smooth muscle actin and desmin (Sigma Aldrich, St. Louis, MO); monoclonal anti-SMemb and polyclonal anti-FAK (Abcam, Cambridge, MA); polyclonal anti-paxillin, monoclonal anti-vinculin and monoclonal anti-ED-A fibronectin (Millipore, Billerica, MA); monoclonal anti-tensin (BD Biosciences, Mississauga, ON); and anti-pro-collagen (sp1.D8) (Hybridoma bank, University of Iowa).
Cell culture reagents were purchased from Gibco BRL unless otherwise specified. Goat anti-rabbit HRP-linked secondary antibodies were purchased from Cell Signaling (New England Biolabs Ltd., Mississauga, Ontario). Biotinylated secondary antibodies and streptavidin FITC were from Amersham-Pharmacia (Baie d'Urfe, QC). The calcium chelator BAPTA was purchased from Sigma. HEPES buffered saline (HBS pH 7.0) was prepared containing 25 ml of 200 mM HEPES, 11.8 ml of 100 mM NaOH, and distilled H2O to 100 ml, then sterilized by passage through a 0.22 μM filter. GdCl3 and other laboratory grade reagents were purchased from Sigma-Aldrich Canada Ltd. (Oakville, Ontario).
Quantified densitometric analysis of each band was done using a computer program (Quantity One 1-D Analysis Software) connected to a scanner (GS-800 Calibrated Densitometer). Within the program, the background subtraction was set to local and the regression method was linear. Data were normalized using GAPDH as loading control. All values were expressed as mean ± SEM. Control and experimental means were compared using analysis of variance followed by either Dunnett's or Bonferonni's post hoc analysis with a significance of * P < 0.05 or higher. P0 to P3 cell extracts were run in triplicate in a single gel to compare each group using a single blot. To generate final figures, representative bands were taken from different lanes along with matching controls from the host blot. Each “n” represents cells taken from a unique animal.
We thank Stephen C. Jones, who helped initiate some of the protocols used in the current study and Robert R. Fandrich for assistance in cryosectioning and immunofluoresence of frozen cardiac tissues. We also thank Jared Davies, Eric Bissonnette, and Vanessa Hunzinger for their practical assistance. This work was carried out with a grant-in-aid from the Heart and Stroke Foundation of Canada (I.M.C.D.), and the CIHR (I.M.C.D. and E.K.). Personnel support was from the CIHR/IMPACT Program (D.H.F.), a CIHR/MHRC studentship (R.H.C.), and studentships from Institute of Cardiovascular Sciences, St. Boniface General Hospital and the St. Boniface General Hospital and Research Foundation (A.D. and K.M.B.). J.J.S. was supported by a studentship from NSERC. The authors have no current or past ties to vendors or companies linked to the commercial reagents used in this investigation, and have no other conflicts to disclose. All experimental protocols for animal studies were approved by the Animal Care Committee of the University of Manitoba, Canada, following guidelines established by the Canadian Institutes of Health Research and the Canadian Council of Animal Care (2010).
- 2006. Smad3 deficiency alters key structural elements of the extracellular matrix and mechanotransduction of wound closure. Proc Natl Acad Sci U S A 103: 9250–9255. , , , , , , , , .
- 1994. Dependence of collagen remodelling on alpha-smooth muscle actin expression by fibroblasts. J Cell Physiol 159: 161–175. , .
- 2003. Noncultured, autologous, skeletal muscle cells can successfully engraft into ovine myocardium. Circulation 107: 3088–3092. , , , , , , , .
- 2007. Essential role of Smad3 in infarct healing and in the pathogenesis of cardiac remodeling. Circulation 116: 2127–2138. , , , , , , , .
- 2002. Diversity, topographic differentiation, and positional memory in human fibroblasts. Proc Natl Acad Sci U S A 99: 12877–12882. , , , , , , .
- 1995. Collagen remodeling after myocardial infarction in the rat heart. Am J Pathol 147: 325–338. , , , .
- 2006. Fetal and adult fibroblasts have similar TGF-beta-mediated, Smad-dependent signaling pathways. Plast Reconstr Surg 117: 2277–2283. , , , .
- 1993. Transforming growth factor-beta 1 induces alpha-smooth muscle actin expression in granulation tissue myofibroblasts and in quiescent and growing cultured fibroblasts. J Cell Biol 122: 103–111. , , , .
- 1990. Nitrendipine binding in congestive heart failure due to myocardial infarction. Circ Res 66: 782–788. , , .
- 1996. Fibroblast growth factor-2 decreases metabolic coupling and stimulates phosphorylation as well as masking of connexin43 epitopes in cardiac myocytes. Circ Res 79: 647–658. , , , ,
- 2000. The epsilon subtype of protein kinase C is required for cardiomyocyte connexin-43 phosphorylation. Circ Res 86: 293–301. , , .
- 1988. Collagen chain mRNAs in isolated heart cells from young and adult rats. J Mol Cell Cardiol 20: 267–276. , , , , , , .
- 2001a. The fibronexus in reactive and tumoral myofibroblasts: further characterisation by electron microscopy. Histol Histopathol 16: 57–70. .
- 2001b. The myofibroblast: an assessment of controversial issues and a definition useful in diagnosis and research. Ultrastruct Pathol 25: 39–50. .
- 2000. Myofibroblasts in reperfused myocardial infarcts express the embryonic form of smooth muscle myosin heavy chain (SMemb). Cardiovasc Res 48: 89–100. , , .
- 2003. Induction of protein synthesis in cardiac fibroblasts by cardiotrophin-1: integration of multiple signaling pathways. Cardiovasc Res 60: 365–375. , , , .
- 2005. Emerging evidence for the role of cardiotrophin-1 in cardiac repair in the infarcted heart. Cardiovasc Res 65: 782–792. , , , , .
- 2001. Transmembrane crosstalk between the extracellular matrix--cytoskeleton crosstalk. Nat Rev Mol Cell Biol 2: 793–805. , , , .
- 2006. FAK-dependent regulation of myofibroblast differentiation. FASEB J 20: 1006–1008. , , , , , .
- 1994. Fibroblasts, myofibroblasts, and wound contraction. J Cell Biol 124: 401–404. .
- 2000. Interaction between angiotensin II and Smad proteins in fibroblasts in failing heart and in vitro. Am J Physiol 279: H3020–H3030. , , , , .
- 2001. Alpha-smooth muscle actin expression upregulates fibroblast contractile activity. Mol Biol Cell 12: 2730–2741. , , , , .
- 2009. Basic fibroblast growth factor induces down-regulation of alpha-smooth muscle actin and reduction of myofibroblast areas in open skin wounds. Wound Repair Regen 17: 617–625. , , , , , , , .
- 2000. Chemotaxis of primitive hematopoietic cells in response to stromal cell-derived factor-1. J Clin Invest 105: 101–111. , , , .
- 2006. Fibroblasts in cancer. Nat Rev Cancer 6: 392–401. , .
- 1991. Cardiac myofibroblasts express alpha smooth muscle actin during right ventricular pressure overload in the rabbit. Am J Pathol 139: 207–216. , , , , .
- 2003. Transforming growth factor-beta 1-mediated collagen gel contraction by cardiac fibroblasts. J Renin Angiotensin Aldosterone Syst 4: 113–118. , , .
- 1996. Myofibroblasts differentiate from fibroblasts when plated at low density. Proc Natl Acad Sci U S A 93: 4219–4223. , , , , .
- 2004. An exploration of two opposing theories of wound contraction. J Wound Care 13: 138–40. .
- 2000. Fibroblast growth factor-2. Int J Biochem Cell Biol 32: 263–267. , , .
- 2008. Perfusion-decellularized matrix: using nature's platform to engineer a bioartificial heart. Nat Med 14: 213–221. , , , , , , .
- 1999. Expression of Gi-2 alpha and Gs alpha in myofibroblasts localized to the infarct scar in heart failure due to myocardial infarction. Cardiovasc Res 41: 575–585. , , , , , .
- 2002. Stimulation of collagen production by transforming growth factor-beta1 during differentiation of cardiac fibroblasts to myofibroblasts. Hypertension 39: 258–263. , , .
- 2010. Non-fibrillar collagens: key mediators of post-infarction cardiac remodeling? J Mol Cell Cardiol 48: 530–537. , .
- 1985. Measurement of protein using bicinchoninic acid. Anal Biochem 150: 76–85. , , , , , , , , , .
- 2008. MicroRNA-21 contributes to myocardial disease by stimulating MAP kinase signalling in fibroblasts. Nature 456: 980–984. , , , , , , , , , , , , , , , , , , , , , , .
- 2002. Myofibroblasts and mechano-regulation of connective tissue remodelling. Nat Rev Mol Cell Biol 3: 349–363. , , , , .
- 2005. Regulation of alpha-smooth muscle actin expression in granulation tissue myofibroblasts is dependent on the intronic CArG element and the transforming growth factor-beta1 control element. Am J Pathol 166: 1343–1351. , , , .
- 1999. Suppression of transforming growth factor-beta isoforms, TGF-beta receptor type II, and myofibroblast differentiation in cultured human corneal and limbal fibroblasts by amniotic membrane matrix. J Cell Physiol 179: 325–335. , , .
- 2000. Paxillin interactions. J Cell Sci 113( pt 23): 4139–4140. .
- 2007. Regulation of collagen synthesis by inhibitory Smad7 in cardiac myofibroblasts. Am J Physiol Heart Circ Physiol 293: H1282–H1290. , , , , , , .
- 2003. Mechanical force regulation of myofibroblast differentiation in cardiac fibroblasts. Am J Physiol Heart Circ Physiol 285: H1871–H1881. , , , .
- 1997. Cardiac interstitium. In: Poole-WilsonPA, ColucciWS, MassieBM, ChatterjeeK, CoatsAJS, editors. Heart failure. New York: Churchill Livingstone. p 13–31. .
- 2007. Myofibroblast contraction activates latent TGF-beta1 from the extracellular matrix. J Cell Biol 179: 1311–1323. , , , .
- 2004. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study. Lancet 364: 937–952. , , , , , , , , , , .