Matrix ProteineCSpecific Regulation of Cx43 Expression in Cardiac Myocytes Subjected to Mechanical Load
http://www.100md.com
循环研究杂志 2005年第3期
the Departments of Pathology and Medicine and the Center for Cardiovascular Research, Washington University School of Medicine, St Louis, Mo.
Abstract
To elucidate mechanisms responsible for mechanotransduction in the heart and define the effects of remodeling of the extracellular matrix, we cultured neonatal rat ventricular myocytes on native type I collagen, fibronectin, or denatured collagen and subjected them to uniaxial, pulsatile stretch. Changes in expression of the cardiac gap junction protein, Cx43, were measured by confocal microscopy and immunoblotting. Cells grown on fibronectin or denatured collagen exhibited significantly greater Cx43 expression than cells grown on native collagen. Stretch induced a 2-fold increase in Cx43 expression in cells grown on native collagen but no increase in cells grown on fibronectin or denatured collagen. Incubation of cells on native collagen with a peptide containing the arginine-glycine-aspartate (RGD) motif upregulated Cx43 expression equivalent to that induced by stretch. Nonselective activation of integrin signaling with MnCl2 also upregulated Cx43 expression in cells grown on native collagen. This effect was blocked completely by pretreatment with anti-1 integrin antibody but not by anti-3 integrin antibody. Stretch led to a marked increase in 1 integrin immunofluorescent signal in cells grown on native collagen but not in cells grown on fibronectin or denatured collagen. Stretch-induced upregulation of Cx43 was also blocked by anti-1 integrin antibody. Thus, matrix protein-myocyte interactions regulate Cx43 expression via 1 integrin signaling initiated by mechanical stimulation in cells grown on native type I collagen, or by RGD-integrin signaling independent of mechanical stress in cells grown on fibronectin or denatured collagen. Changes in the composition of the extracellular matrix may affect electrical coupling in cardiac myocytes.
Key Words: cell culture connexin43 expression extracellular matrix mechanical stretch mechanotransduction
Introduction
Because of the unique mechanical function of the heart, interactions between cardiac myocytes and the extracellular matrix undoubtedly play an important role in regulating fundamental biological and pathobiological processes.1,2 Whereas changes in the structure and composition of the extracellular matrix may be maladaptive in end-stage heart disease, early responses of cardiac myocytes to matrix remodeling could fulfill adaptive purposes. Previously, we reported that expression of the cardiac gap junction protein, connexin43 (Cx43), is increased by 2-fold and conduction of electrical impulses is enhanced when neonatal rat ventricular myocytes grown on native type 1 collagen are subjected to a mechanical load (pulsatile stretch to 110% of resting length at 3 Hz) for only 1 hour.3 Stretch-induced upregulation of Cx43 is mediated at least in part by vascular endothelial growth factor (VEGF), which is secreted in response to stretch and acts in an autocrine manner to enhance Cx43 expression.4 These observations indicate that stretch can lead to salutary changes in cardiac myocytes presumably via activation of intracellular signaling pathways initiated by myocyte-matrix interactions. This in vitro stretch system is well suited for analysis of cardiac myocyte-matrix interactions. It consists of a monolayer of virtually pure cardiac myocytes that can be grown on chemically defined matrix proteins and subjected to controlled mechanical load. The rapid and marked increase in Cx43 expression provides a sensitive and convenient read-out to monitor the effects of altered myocyte-matrix protein interactions on gene expression.
Mechanosensitive integrin receptors have been shown to directly connect the extracellular matrix to the actin cytoskeleton of cardiac myocytes.1,2 Moreover, changes in traction force initiate intracellular signal transduction cascades via "outside-in" integrin signaling.1,2 In the present study, we used the in vitro stretch system to characterize the effects of different matrix proteins on stretch-induced changes in cardiac myocyte gene expression using Cx43 as the read-out, and to elucidate mechanisms responsible for outside-in signaling mediated by integrins. We compared effects in cells grown on native type I collagen, the predominant matrix protein of the normal heart, and fibronectin or denatured type I collagen to model changes known to occur in the diseased heart.5eC9 We observed marked disparities in responses to stretch and VEGF in myocytes grown on native collagen compared with myocytes grown on fibronectin and denatured collagen in which signaling mediated by interactions between the arginine-glycine-aspartate (RGD) motif and 1 integrins played a major role.
Materials and Methods
Neonatal Rat Ventricular Myocyte Cultures
Primary cultures were prepared from ventricles of 1-day-old outbred Wistar rat pups (Charles River, Indianapolis, Ind) as described previously.3,4 Cells were plated at a density of 2.4x105 cells/cm2 on deformable silicone membranes that had been coated with native type I collagen (200 e/mL, Sigma), fibronectin (200 e/mL, Sigma), or heat-denatured (60°C for 30 minutes) type I collagen (200 e/mL). Cells were grown in M199 medium in Hanks balanced salt solution at 37°C in an atmosphere of room air supplemented with 1% CO2 for 4 days before experimentation. All experimental protocols were approved by the Washington University Animal Studies Committee.
Stretch Protocols and Incubation of Cells With VEGF, RGD Peptide, MnCl2, and Anti-1 and Anti-3 Integrin Antibodies
Monolayers of ventricular myocytes that had been grown in culture for 4 days were subjected to uniaxial, pulsatile stretch using a custom-designed apparatus described in detail in a previous study.3 Cells were transferred to serum-free medium (1:1, Ham’s F-12:DMEM, Fisher Life Sciences) and stretched to 110% of resting length at a frequency of 3 Hz for 1 hour. In selected experiments, unstretched myocyte cultures were incubated for 1 hour in serum-free medium containing exogenous recombinant human VEGF165 (the 165 amino acid isoform of human VEGF; 100 ng/mL, R&D Systems), a cyclic RGD peptide (GRDGSP) or an inactive RGE peptide (GRGESP) (0.5 mmol/L each, Washington University Protein Chemistry Laboratory), or MnCl2 (1 mmol/L, Sigma). In other experiments, cultures were pretreated for 30 minutes in serum-free media containing 50 e/mL monoclonal anti-1 (clone Ha2/5, BD Pharmingen) or anti-3 integrin antibodies (clone 2C9.G2, BD Pharmingen) before being subjected to pulsatile stretch or incubation with other reagents. These antibodies have been shown in previous studies to block signaling mediated by 1 and 3 integrins, respectively.10,11 Control cultures (not stretched or exposed to exogenous reagents) were also incubated in serum-free medium for 1 hour. At the conclusion of each experimental protocol, cells were either scraped from membranes and lysed for immunoblot analysis or fixed for confocal immunofluorescence microscopy.
Immunoblot Assay of Cx43 Expression
Cell cultures were washed once in cold PBS and then scraped from silicone membranes in a low ionic strength buffer containing protease inhibitors (NaHCO3,1 mmol/L; EDTA, 5 mmol/L; EGTA, 1 mmol/L; leupeptin, 1 eol/L; pepstatin, 1 eol/L; aprotinin, 0.1 eol/L; benzamidine, 1 mmol/L; iodoacetamide, 1 mmol/L; phenylmethylsulfonylfluoride, 1 mmol/L). After centrifugation at 3000g for 6 minutes at 4°C, the pellet was resuspended in the same buffer and stored at eC70°C. Aliquots containing 7 e of total protein were analyzed by SDS polyacrylamide gel electrophoresis, and Cx43 signal was quantified by densitometry as described in detail in previous studies.12
Quantitative Confocal Immunofluorescence Microscopy
Cultures were rinsed in serum-free medium and then fixed in 4% paraformaldehyde. To measure the amount of Cx43 immunoreactive signal located in gap junctions, cells were incubated overnight with a polyclonal rabbit anti-Cx43 antibody (diluted 1:400; product no. 71-0700, Zymed) and then with Cy3-conjugated goat anti-rabbit IgG (diluted 1:400, Jackson Immunoresearch) before being examined by laser scanning confocal microscopy. In selected experiments, expression of N-cadherin was analyzed using a rabbit polyclonal antibody that recognizes a common epitope in the N-cadherins (product no. C3678, Sigma). The area occupied by specific Cx43 or N-cadherin immunoreactive signal in cell-cell junctions was quantified by digital image processing and expressed as a percentage of total myocyte area using established, validated methods that have been described previously in detail.13 In each culture, three fields were randomly selected for quantitative analysis. Each field was 26 400 e2 in area and included profiles of 30 to 50 cells. A mean value was calculated for each culture by averaging the results from the three fields. At least three individual cultures were analyzed for each condition (n3 experiments). The amount and distribution of 1 integrin signal was also characterized by immunofluorescence microscopy in paraformaldehyde-fixed cells, which were incubated overnight with an Armenian hamster anti-1 integrin antibody (diluted 1:1000, BD Pharmingen) and then with Cy3-conjugated goat anti-Armenian hamster IgG (diluted 1:400, Jackson Immunoresearch).
Dye-Transfer Assay
Cells grown on silicone membranes coated with type I collagen or fibronectin (n=3 membranes for each) were rinsed in PBS and then incubated with Lucifer yellow (1%, Sigma). A small glass-cutting wheel was used to make a clean cut across the myocyte monolayer to permit the membrane-impermeant dye to enter a linear array of broken cells (scrape-loading of dye). After being incubated for 2 minutes at 37°C, the preparations were rinsed three times with PBS, fixed in 4% paraformaldehyde, mounted on glass slides, and photographed using a fluorescence microscope and Nikon Coolpix 995 camera. Digital images were analyzed off-line using ImageJ software (NIH, http://rsb.info.nih.gov/ij/). Three fields from each culture were analyzed. An area of dye transfer adjacent to the scrape line was obtained from each field and divided by the length over which the area was measured, to yield an average linear value of dye-spread expressed in micrometers (e).
Statistical Analysis
Data are expressed as mean±SD. Differences between groups were analyzed with ANOVA and Fisher protected least significant difference test. A value of P<0.05 was considered sufficient to reject the null hypothesis.
Results
Myocyte-Matrix Protein Interactions Affect Cx43 Expression at Gap Junctions
To determine whether the chemical composition of the extracellular matrix influences Cx43 expression in cardiac myocytes grown under basal conditions (without stretch), cultures were plated on either native type I collagen or fibronectin, grown for 4 days, and Cx43 expression at cell-cell junctions was measured by quantitative confocal microscopy. As shown in representative confocal immunofluorescent images and quantitative confocal analysis of three independent sets of cultures (Figure 1), the amount of Cx43 immunoreactive signal at sites of intercellular junctions was 2-fold greater in myocyte cultures grown on fibronectin than on collagen (2.19%±0.61 versus 1.05%±0.13, respectively; P<0.003). Additional cells grown on native collagen or fibronectin were subjected to pulsatile stretch or incubated with exogenous VEGF for 1 hour. As previously shown,3,4 myocytes plated on collagen showed a significant increase in the amount of Cx43 immunoreactive signal at intercellular junctions in response to pulsatile stretch (from 1.05±0.13% to 2.08±0.74%; P<0.04) and after being incubated with VEGF for 1 hour (from 1.05±0.13% to 2.48±0.60%; P<0.02) (Figure 1). However, neither stretch nor VEGF increased the amount of Cx43 immunoreactive signal in myocytes plated on fibronectin (Figure 1). The amount of Cx43 expression in cells grown on fibronectin was roughly equal to the level observed after stretch or incubation with VEGF in cells grown on native type I collagen. No apparent change in cell shape was noted in cells grown on fibronectin or subjected to stretch.
Previously, we showed that enhanced Cx43 expression in response to stretch in cells grown on collagen was associated with increased coupling.3,4 To determine whether increased Cx43 expression in nonstretched cells grown on fibronectin was also associated with enhanced coupling, we scrape-loaded cells with the membrane-impermeant dye, Lucifer yellow, and measured the extent to which dye passed to adjacent cells through gap junctions. As shown in Figure 1, dye traveled 71% further in nonstretched cells grown on fibronectin (229±35 e) than in nonstretched cells grown on collagen (134±41 e; n=3 for each, P<0.04).
Integrin-Matrix Protein Interactions Alter Cx43 Expression at Gap Junctions
Fibronectin contains abundant RGD cell adhesive motifs.13,14 Furthermore, myocardial injury during ischemia or infarction leads to exposure of previously cryptic RGD motifs in native collagen in the myocardial interstitium.8,9,14eC16 To determine whether denaturation of collagen, which is known to expose RGD motifs, would produce changes in Cx43 expression similar to those seen in cells grown on fibronectin, myocytes were grown on either native collagen or collagen subjected to heat-denaturation. As shown in representative confocal immunofluorescent images and quantitative confocal analysis from three independent experiments (Figure 2), the amount of Cx43 immunoreactive signal at sites of intercellular junctions was 1.6-fold greater in myocytes grown on heat-denatured collagen under basal conditions than in myocytes grown under identical conditions on native collagen (1.96%±0.63 versus 1.19%±0.28, respectively; P<0.03). As was observed in cells grown on fibronectin, myocytes plated on heat-denatured collagen showed no significant increase in Cx43 signal in response to 1 hour of pulsatile stretch or incubation with exogenous VEGF (Figure 2). Immunoblotting demonstrated an increase in the total amount of Cx43 in response to stretch in myocytes grown on native collagen (Figure 2), indicating that enhanced Cx43 signal at cell-cell junctions seen by confocal microscopy reflected an increase in the total Cx43 content within cardiac myocytes rather than only translocation from intracellular to junctional pools. A similarly increased level of Cx43 expression was also seen in nonstretched cells grown on denatured collagen, and there was no further increase after stretch (Figure 2). No apparent changes in Cx43 phosphorylation were observed, at least as judged by the relative intensities of the various Cx43 bands on immunoblots. No changes in cell shape were noted either.
Previously, we found that expression of the intercellular adhesion molecule N-cadherin was increased in response to stretch in cells grown on native type I collagen.3 To determine whether expression of this mechanical junction protein was also affected by different matrix proteins, cells grown on fibronectin and denatured collagen were subjected to stretch and N-cadherin expression was measured by confocal microscopy. As shown in Figure 3, N-cadherin expression was increased in nonstretched cells grown on fibronectin or denatured collagen compared with nonstretched cells grown on native collagen. Stretch increased N-cadherin expression by >3-fold in cells on native collagen but had no effect in cells grown on fibronectin or denatured collagen. These results suggest that expression of electrical and mechanical junction proteins is coordinately regulated.
Results in cells grown on fibronectin or denatured collagen suggest that integrin signaling initiated by RGD motifs upregulates Cx43 expression and increases the amount of Cx43 in gap junctions independent of stretch and VEGF. To further determine the effect of the RGD motif on Cx43 expression, myocytes plated on native type I collagen were incubated with a cyclic RGD peptide (GRGDSP). Although this peptide can be used to block integrin signaling, we observed a marked increase in Cx43 expression. As shown in representative confocal images and by quantitative confocal analysis (Figure 4), the amount of Cx43 immunoreactive signal at intercellular junctions was >2-fold greater (2.57±0.60% versus 1.10±0.06%, respectively; P<0.002) in myocytes incubated with the RGD peptide than in control cultures not treated with peptide. Incubation of cells with an RGE peptide had no effect (data not shown). Because cellular expression and recruitment of specific integrin heterodimers occurs with RGD exposure, these findings suggest that activation of integrin heterodimers alters myocyte expression of Cx43.
Generalized Activation of Integrins Increases Cx43 Expression at Gap Junctions
To further investigate the role of integrin signaling in regulating Cx43 expression, we exposed cells to MnCl2. The divalent cation Mn2+ stabilizes an open, extended integrin conformation promoting constitutive, high-affinity ligand binding.17 To determine whether generalized integrin activation affects Cx43 expression at gap junctions, myocytes grown on native type I collagen or heat-denatured collagen were exposed to 1 mmol/L MnCl2 for 1 hour. No change in pH was detected in the media of cells incubated with MnCl2. The amount of Cx43 immunoreactive signal at intercellular junctions was significantly increased in myocytes on native collagen that were incubated with MnCl2 (from 1.28±0.14% to 2.11±0.20%; P<0.001) but not in cells grown on heat-denatured collagen (Figure 5). Immunoblotting demonstrated an increase in the total content of Cx43 in myocytes incubated with MnCl2 (Figure 5) Thus, generalized activation of integrins upregulates Cx43 expression and increases the amount of Cx43 in gap junctions independent of mechanical perturbation.
Role for 1 Integrin Signaling in Mechanoelectrical Coupling
1 Integrins have been shown to play an essential role as mechanotransducers in the heart.1,2,18eC21 Although cardiac myocytes express both 1 and 3 integrin subunits,1,2 native collagen binds only to integrin heterodimers containing the 1 integrin subunit.22 To further elucidate the respective roles of 1 and 3 integrins in regulating Cx43 expression, myocytes grown on native collagen were exposed to MnCl2 in the presence or absence of monoclonal antibodies against 1 or 3 integrin subunits. As previously demonstrated, addition of MnCl2 significantly increased Cx43 expression at gap junctions (2.21±0.42% versus 1.05±0.21%, respectively; P<0.04) compared with controls (Figure 6). Pretreatment with anti-1 integrin antibody completely blocked the increase in Cx43 expression at gap junctions caused by MnCl2 (1.04±0.04% versus 1.05±0.21%, respectively; P=NS) (Figure 6). In contrast, anti-3 integrin antibody had no effect on the degree of upregulation of Cx43 expression induced by MnCl2 (Figure 6). Immunoblotting showed that pretreatment with anti-1 integrin antibody apparently decreased total Cx43 content in cells incubated with MnCl2 despite the fact that Cx43 signal in junctions was not different than in control cells (Figure 6). The significance of this specific observation is unclear but overall, the results of this experiment indicate that generalized activation of integrin signaling by MnCl2 upregulates Cx43 expression by a 1 integrin-specific mechanism.
To further characterize the role of 1 integrins in mechanotransduction, myocytes grown on native collagen were subjected to pulsatile stretch for 1 hour and the amount and distribution of 1 integrin immunofluorescent signal was analyzed by confocal microscopy. As shown in Figure 7, the amount of 1 immunoreactive signal was markedly increased in response to stretch. However, no apparent increase in 1 integrin signal was noted in unstretched cells grown on fibronectin or denatured collagen, or in cells on native collagen that were incubated with Mn2+. Finally, to determine whether 1 integrin signaling mediates stretch-induced upregulation of Cx43 expression, cells grown on native collagen were stretched in the presence or absence of anti-1 integrin antibody. As previously shown, stretch led to significant upregulation of Cx43 expression (from 1.05±0.07% to 1.91±0.18%, n=4; P<0.001) (Figure 8). However, pretreatment with anti-1 integrin antibody blocked stretch-induced upregulation of Cx43 expression (Figure 8). Addition of exogenous VEGF to nonstretched cells that had been pretreated with anti-1 integrin antibody upregulated Cx43 expression (to 1.83%±0.40, n=4; P<0.02 versus control cells) (Figure 8). Thus, blockade of integrin signaling does not prevent cells from responding to VEGF.
Discussion
The present study was designed to elucidate the effects of specific matrix proteins in regulating cardiac myocyte gene expression induced by mechanical load. We compared the effects of short-term pulsatile stretch on ventricular myocytes grown on native type I collagen, the major component of the extracellular matrix in the normal heart, and myocytes grown on fibronectin and denatured collagen, both of which accumulate in the myocardial interstitium under selected pathophysiological conditions. We focused specifically on effects of mechanical load on expression of Cx43, the major cardiac gap junction channel protein, in ventricular myocytes grown on the different matrix protein. In previous studies,3 we observed that Cx43 expression increased by 2-fold after 1 hour of pulsatile stretch in ventricular myocytes grown on type I collagen. We also showed previously that VEGF, secreted during stretch, acted through an autocrine mechanism to upregulate Cx43 expression, and that the stretch effect could be mimicked by incubating nonstretched cells with exogenous VEGF.4 In the present study, we found that when ventricular myocytes were grown on either fibronectin or denatured collagen, they expressed increased amounts of Cx43 compared with cells grown on type I collagen. Cx43 levels in nonstretched cells on fibronectin or denatured collagen were roughly equivalent to the levels seen after stretch in cells grown on native type I collagen. Moreover, neither stretch nor exogenous VEGF increased Cx43 expression in cells grown on fibronectin or denatured collagen. Similar results were observed for the intercellular adhesion molecule, N-cadherin. These observations suggest that signaling pathways activated by stretch or VEGF in cells grown on native collagen are constitutively activated in cells grown on fibronectin or denatured collagen such that no further increase in Cx43 expression is caused by stretch or VEGF.
In selected experiments, we observed that increased Cx43 expression at cell-cell junctions was accompanied by an increase in total Cx43 content measured by immunoblotting. This confirms previous findings.4 We have also shown previously that increased Cx43 expression induced by stretch3,4 or by exposure to cAMP23 in vitro leads to enhanced coupling as demonstrated by increased conduction velocity. In the present study, we showed that increased Cx43 expression in cells grown on fibronectin was associated with enhanced junctional coupling as demonstrated using a standard dye-transfer assay. These results provide further evidence that upregulation of Cx43 expression at cell-cell junctions, measured by confocal microscopy, reflects an increase in functional gap junction channels.
Because both fibronectin and denatured collagen, but not native collagen, interact with integrins via RGD motifs,8,9,14eC16 we reasoned that activation of integrin signaling by RGD was responsible for enhanced Cx43 expression under basal conditions in cells grown on fibronectin or denatured collagen. Indeed, we could mimic the stretch effect by incubating cells on native collagen with a cyclic RGD peptide to specifically activate integrins dependent on RGD signaling, or by activating integrins nonspecifically with MnCl2. No further increase in Cx43 expression was achieved by exposing cells on fibronectin or denatured collagen to MnCl2. Thus, integrin signaling responsible for upregulating Cx43 expression appears to be maximal in cells grown on fibronectin or denatured collagen.
Finally, we showed that enhanced Cx43 expression induced by incubating cells grown on native collagen with MnCl2 could be blocked completely by an anti-1 integrin antibody but not an anti-3 integrin antibody. Moreover, stretch-induced upregulation of Cx43 expression was also blocked by anti-1 integrin antibody in cells grown on native collagen. These results indicate that signaling activated by mechanical load in cardiac myocytes is mediated by 1 integrins. The fact that the effects of Mn2+ was blocked by anti-1 integrin antibody also suggests that Mn2+ increased Cx43 expression by activating integrin signaling rather than by other potential mechanisms.
Several classes of molecules may function as mechanosensors to regulate cardiac myocyte responses to mechanical load. Previous studies have identified only a limited role of stretch-activated ion channels (pacemaker inward, ATP-sensitive K+, and stress sensitive cation channels) in sensing mechanical stimuli and transducing a hypertrophic response in myocytes.24,25 Whereas intercellular adhesion molecules (ICAMs) also activate signaling pathways,26 only the integrins directly connect the extracellular matrix to the actin cytoskeleton, and many studies have demonstrated that transduction of mechanical stress in myocytes into biochemical signals is mediated by integrins.1,2
Type I and III collagens constitute the great majority of the extracellular matrix in the normal heart, but production and secretion of fibronectin by fibroblasts is a key early step in establishment of the matrix in the developing heart.27,28 Fibronectin interacts with collagen, heparin, fibrin, and cell surface integrins to bind diverse matrix components and serve as a scaffold for nascent matrix assembly.15 In the present studies, nearly pure populations of cardiac myocytes were grown on defined matrix proteins to delineate specific matrix-myocyte interactions. Because contamination by fibroblasts is low (<5%),4 the effects of fibronectin and/or other matrix components secreted by fibroblasts in cultures plated on collagen were probably negligible. Similarly, there were no apparent effects of fibronectin that may have been derived from serum in cell culture media.
Various pathological conditions are associated with progressive qualitative and quantitative changes in the composition of the extracellular matrix, which may lead to interstitial fibrosis.27,29,30 Increased levels of mRNAs encoding fibronectin and collagen have been observed during the transition from hypertrophy to failure in spontaneously hypertensive rats and in pressure-overload hypertrophy induced by aortic constriction in rats.5eC7 Similarly, denaturation and matrix metalloproteinaseeCmediated degradation of collagen induced by acute myocardial ischemia have been shown to expose previously cryptic RGD motifs within interstitial type I collagen, and this may be followed by increased expression of fibronectin.8,9 Although the defined in vitro conditions used in the present studies cannot precisely reflect changes in the matrix in cardiac disease, our observations provide new insights about how matrix remodeling and RGD-mediated activation of integrins might activate signaling cascades that promote adaptive responses to injury.
Enhanced Cx43 expression is known to occur during early, compensatory phases of cardiac hypertrophy,31 but the potential role of altered myocyte-matrix interactions in this context are unknown.32 Remodeling of gap junctions is also well documented in regions bordering acute myocardial infarcts. The appearance of RGD sites within discrete areas of acute injury may help regulate the repair process and/or initiate adaptive responses in viable but injured cells adjacent to the zone of necrosis. Alternatively, nonuniform, heterogeneous remodeling of the extracellular matrix might contribute to development of discrete areas of differential Cx43 expression and account for enhanced anisotropic conduction and formation of functional lines of block, which could promote reentrant arrhythmias.33 The present study provides evidence for a potential new mechanism linking remodeling of the extracellular matrix to changes in electrical coupling of cardiac myocytes.
The 1 integrin is the predominant -subunit found in the myocardium and has been implicated in mediating the cardiac hypertrophic response.18eC22 A crucial role for 1 integrin signaling in sensing shear stress in endothelial cells has also been well established.34 Our findings that 1 integrin expression increases with pulsatile stretch (a result that has been demonstrated previously with static stretch)35 and that 1 integrin signaling is responsible for upregulation of Cx43 induced by MnCl2 provide additional evidence for a primary role of 1 integrins as mechanotransducers in cardiac myocytes. Our observations suggest further that altered expression of Cx43 and remodeling of intercellular electrical junctions could play a role in the pathogenesis of the lethal phenotype observed in 1D integrin knockout mice after transaortic constriction.20
In conclusion, we have shown that matrix protein-myocyte interactions can regulate Cx43 expression in gap junctions via 1 integrin signaling initiated by mechanical stimulation in cells grown on native type I collagen, or by RGD-integrin signaling independent of mechanical stress in cells grown on fibronectin or denatured collagen. The composition of the extracellular matrix in physiological or pathophysiological states may, therefore, affect the degree of electrical coupling in cardiac myocytes.
Acknowledgments
This work was supported by NIH grants HL50598 and HL74595. We thank Dr Steven Teitelbaum for helpful discussions and critical review of the manuscript.
References
Katsumi A, Orr AW, Tzima E, Schwartz MA. Integrins in mechanotransduction. J Biol Chem. 2004; 279: 12001eC12004.
Ross RS, Borg TK. Integrins and the myocardium. Circ Res. 2001; 88: 1112eC1119.
Zhuang J, Yamada KA, Saffitz JE, Kleeber AG. Pulsatile stretch remodels cell-to-cell communication in cultured myocytes. Circ Res. 2000; 87: 316eC322.
Pimentel RC, Yamada KA, Kleeber AG, Saffitz JE. Autocrine regulation of myocyte Cx43 expression by VEGF. Circ Res. 2002; 90: 671eC677.
Villareal FJ, Dillman WH. Cardiac hypertrophy-induced changes in mRNA levels for TGF-1, fibronectin, and collagen. Am J Physiol. 1992; 31: H1861eCH1866.
Boluyt MO, O’Neill L, Meredith AL, Bing OH, Brooks WW, Conrad CH, Crow MT, Lakatta EG. Alterations in cardiac gene expression during the transition from stable hypertrophy to heart failure: marked upregulation of genes encoding extracellular matrix proteins. Circ Res. 1994; 75: 23eC32.
Boluyt MO, Bing OH. Matrix gene expression and decompensated heart failure: the aged SHR model. Cardiovasc Res. 2000; 46: 239eC249.
Cannon RO III, Butany JW, McManus BM, Speir E, Kravitz AB, Bolli R, Ferrans VJ. Early degradation of collagen after acute myocardial infarction in the rat. Am J Cardiol. 1983; 52: 390eC395.
Davis GE, Bayless KJ, Davis MJ, Meininger GA. Regulation of tissue injury responses by the exposure of matricryptic sites within extracellular matrix molecules. Am J Pathol. 2000; 156: 1489eC1498.
Schultz JF, Armant DR. Beta 1- and beta 3-class integrins mediate fibronectin binding activity at the surface of developing mouse peri-implantation blastocysts: regulation by ligand-induced mobilization of stored receptor. J Biol Chem. 1995; 270: 11522eC11531.
Ogawa E, Saito Y, Harada M, Kamitani S, Kuwahara K, Miyamoto Y, Ishikawa M, Hamanaka I, Kajiyama N, Takahashi N, Nakagawa O, Masuda I, Kishimoto I, Nakao K. Outside-in signaling of fibronectin stimulates cardiomyocyte hypertrophy in cultured neonatal rat ventricular myocytes. J Mol Cell Cardiol. 2000; 32: 765eC776.
Johnson CM, Green KG, Kanter EM, Bou-Abboud E, Saffitz JE, Yamada KA. Voltage-gated Na2+ channel activity and connexin expression in Cx43-deficient myocytes. J Cardiovasc Electrophysiol. 1999; 10: 1390eC1401.
Saffitz JE, Green KG, Kraft WJ, Schechtman KB, Yamada KA. Effects of diminished expression of connexin43 on gap junction number and size in ventricular myocardium. Am J Physiol Heart Circ Physiol. 2000; 278: H1662eCH1670.
Yamamoto K, Yamamoto M. Cell adhesion receptors for native and denatured type I collagens and fibronectin in rabbit arterial smooth muscle cells in culture. Exp Cell Res. 1994; 214: 258eC263.
Wierzbicka-Patynowski I, Schwarzbauer JE. The ins and outs of fibronectin matrix assembly. J Cell Sci. 2003; 116: 3269eC3276.
Davis GE. Affinity of integrins for damaged extracellular matrix: alpha v beta 3 binds to denatured collagen type I through RGD sites. Biochem Biophys Res Commun. 1992; 182: 1025eC1031.
Takagi J, Petre BM, Walz T, Springer TA. Global conformational rearrangements I integrin extracellular domains in outside-in and inside-out signaling. Cell. 2002; 110: 599eC611.
Ross RS, Pham C, Shai SY, Goldhaber JI, Fenczik C, Glembotski CC, Ginsberg MH, Loftus JC. Beta1 integrins participate in the hypertrophic response of rat ventricular myocytes. Circ Res. 1998; 82: 1160eC1172.
Pham CG, Harpf AE, Keller RS, Vu HT, Shai SY, Loftus JC, Ross RS. Striated muscle-specific beta(1D)-integrin and FAK are involved in cardiac myocyte hypertrophic response pathway. Am J Physiol Heart Circ Physiol. 2000; 279: H2916eCH2926.
Shai SY, Harpf AE, Babbitt CJ, Jordan MC, Fishbein MC, Chen J, Omura M, Leil TA, Becker KD, Jiang M, Smith DJ, Cherry SR, Loftus JC, Ross RS. Cardiac myocyte-specific excision of the beta1 integrin gene results in myocardial fibrosis and cardiac failure. Circ Res. 2002; 90: 458eC464.
Brancaccio M, Fratta L, Notte A, Hirsch E, Poulet R, Guazzone S, De Acetis M, Vecchione C, Marino G, Altruda F, Silengo L, Tarone G, Lembo G. Melusin, a muscle-specific integrin beta1-interacting protein, is required to prevent cardiac failure in response to chronic pressure overload. Nat Med. 2003; 9: 68eC75.
Gullberg DE, Lundgren-Akerlund E. Collagen-binding I domain integrins: what do they do Prog Histochem Cytochem. 2002; 37: 3eC54.
Darrow BJ, Fast VG, Kleeber AG, Beyer EC, Saffitz JE. Functional and structural assessment of intercellular communication: increased conduction velocity and enhanced connexin expression in dibutyryl cAMP-treated cultured cardiac myocytes. Circ Res. 1996; 79: 174eC183.
Sadoshima J, Takahashi T, Jahn L, Izumo S. Roles of mechano-sensitive ion channels, cytoskeleton, and contractile activity in stretch-induced immediate-early gene expression and hypertrophy of cardiac myocytes. Proc Natl Acad Sci U S A. 1992; 89: 9905eC9909.
Yamazaki T, Komuro I, Kudoh S, Zou Y, Nagai R, Aikawa R, Uozumi H, Yazaki Y. Role of ion channels and exchangers in mechanical stretch-induced cardiomyocyte hypertrophy. Circ Res. 1998; 82: 430eC437.
Thiery JP. Cell adhesion in development: a complex signaling network. Curr Opin Genet Dev. 2003; 13: 365eC371.
Weber KT, Brilla CG. Pathological hypertrophy and the cardiac interstitium. Circulation. 1991; 83: 1849eC1865.
Raghow R. The role of extracellular matrix in postinflammatory wound healing and fibrosis. FASEB J. 1994; 8: 823eC831.
Lindsey ML, Mann DL, Entman ML, Spinale FG. Extracellular matrix remodeling following myocardial injury. Ann Med. 2003; 35: 316eC326.
Janicki JS, Brower GL, Gardner JD, Chancey AL, Stewart JA Jr. The dynamic interaction between matrix metalloproteinase activity and adverse myocardial remodeling. Heart Fail Rev. 2004; 9: 33eC42.
Kanno S, Saffitz JE. The role of gap junctions in electrical conduction and arrhythmogenesis in the heart. Cardiovasc Pathol. 2001; 10: 169eC177.
Peters NS, Coromilas J, Severs NJ, Wit AL. Disturbed connexin43 gap junction distribution correlates with the location of reentrant circuits in the epicardial border zone of healing canine infarcts that cause ventricular tachycardia. Circulation. 1997; 95: 988eC996.
van Rijen HV, Eckardt D, Degen J, Theis M, Ott T, Willecke K, Jongsma HJ, Opthof T, de Bakker JM. Slow conduction and enhanced anisotropy increase the propensity for ventricular tachyarrhythmias in adult mice with induced deletion of connexin43. Circulation. 2004; 109: 1048eC1055.
Shyy JY, Chien S. Role of integrins in endothelial mechanosensing of shear stress. Circ Res. 2002; 91: 769eC775.
Sharp WW, Simpson DC, Borg TK, Samarel AM, Terracio L. Mechanical forces regulate focal adhesion and costamere assembly in cardiac myocytes. Am J Physiol Heart Circ Physiol. 1997; 273: H546eCH556.(Amit J. Shanker, Kiyomi Y)
Abstract
To elucidate mechanisms responsible for mechanotransduction in the heart and define the effects of remodeling of the extracellular matrix, we cultured neonatal rat ventricular myocytes on native type I collagen, fibronectin, or denatured collagen and subjected them to uniaxial, pulsatile stretch. Changes in expression of the cardiac gap junction protein, Cx43, were measured by confocal microscopy and immunoblotting. Cells grown on fibronectin or denatured collagen exhibited significantly greater Cx43 expression than cells grown on native collagen. Stretch induced a 2-fold increase in Cx43 expression in cells grown on native collagen but no increase in cells grown on fibronectin or denatured collagen. Incubation of cells on native collagen with a peptide containing the arginine-glycine-aspartate (RGD) motif upregulated Cx43 expression equivalent to that induced by stretch. Nonselective activation of integrin signaling with MnCl2 also upregulated Cx43 expression in cells grown on native collagen. This effect was blocked completely by pretreatment with anti-1 integrin antibody but not by anti-3 integrin antibody. Stretch led to a marked increase in 1 integrin immunofluorescent signal in cells grown on native collagen but not in cells grown on fibronectin or denatured collagen. Stretch-induced upregulation of Cx43 was also blocked by anti-1 integrin antibody. Thus, matrix protein-myocyte interactions regulate Cx43 expression via 1 integrin signaling initiated by mechanical stimulation in cells grown on native type I collagen, or by RGD-integrin signaling independent of mechanical stress in cells grown on fibronectin or denatured collagen. Changes in the composition of the extracellular matrix may affect electrical coupling in cardiac myocytes.
Key Words: cell culture connexin43 expression extracellular matrix mechanical stretch mechanotransduction
Introduction
Because of the unique mechanical function of the heart, interactions between cardiac myocytes and the extracellular matrix undoubtedly play an important role in regulating fundamental biological and pathobiological processes.1,2 Whereas changes in the structure and composition of the extracellular matrix may be maladaptive in end-stage heart disease, early responses of cardiac myocytes to matrix remodeling could fulfill adaptive purposes. Previously, we reported that expression of the cardiac gap junction protein, connexin43 (Cx43), is increased by 2-fold and conduction of electrical impulses is enhanced when neonatal rat ventricular myocytes grown on native type 1 collagen are subjected to a mechanical load (pulsatile stretch to 110% of resting length at 3 Hz) for only 1 hour.3 Stretch-induced upregulation of Cx43 is mediated at least in part by vascular endothelial growth factor (VEGF), which is secreted in response to stretch and acts in an autocrine manner to enhance Cx43 expression.4 These observations indicate that stretch can lead to salutary changes in cardiac myocytes presumably via activation of intracellular signaling pathways initiated by myocyte-matrix interactions. This in vitro stretch system is well suited for analysis of cardiac myocyte-matrix interactions. It consists of a monolayer of virtually pure cardiac myocytes that can be grown on chemically defined matrix proteins and subjected to controlled mechanical load. The rapid and marked increase in Cx43 expression provides a sensitive and convenient read-out to monitor the effects of altered myocyte-matrix protein interactions on gene expression.
Mechanosensitive integrin receptors have been shown to directly connect the extracellular matrix to the actin cytoskeleton of cardiac myocytes.1,2 Moreover, changes in traction force initiate intracellular signal transduction cascades via "outside-in" integrin signaling.1,2 In the present study, we used the in vitro stretch system to characterize the effects of different matrix proteins on stretch-induced changes in cardiac myocyte gene expression using Cx43 as the read-out, and to elucidate mechanisms responsible for outside-in signaling mediated by integrins. We compared effects in cells grown on native type I collagen, the predominant matrix protein of the normal heart, and fibronectin or denatured type I collagen to model changes known to occur in the diseased heart.5eC9 We observed marked disparities in responses to stretch and VEGF in myocytes grown on native collagen compared with myocytes grown on fibronectin and denatured collagen in which signaling mediated by interactions between the arginine-glycine-aspartate (RGD) motif and 1 integrins played a major role.
Materials and Methods
Neonatal Rat Ventricular Myocyte Cultures
Primary cultures were prepared from ventricles of 1-day-old outbred Wistar rat pups (Charles River, Indianapolis, Ind) as described previously.3,4 Cells were plated at a density of 2.4x105 cells/cm2 on deformable silicone membranes that had been coated with native type I collagen (200 e/mL, Sigma), fibronectin (200 e/mL, Sigma), or heat-denatured (60°C for 30 minutes) type I collagen (200 e/mL). Cells were grown in M199 medium in Hanks balanced salt solution at 37°C in an atmosphere of room air supplemented with 1% CO2 for 4 days before experimentation. All experimental protocols were approved by the Washington University Animal Studies Committee.
Stretch Protocols and Incubation of Cells With VEGF, RGD Peptide, MnCl2, and Anti-1 and Anti-3 Integrin Antibodies
Monolayers of ventricular myocytes that had been grown in culture for 4 days were subjected to uniaxial, pulsatile stretch using a custom-designed apparatus described in detail in a previous study.3 Cells were transferred to serum-free medium (1:1, Ham’s F-12:DMEM, Fisher Life Sciences) and stretched to 110% of resting length at a frequency of 3 Hz for 1 hour. In selected experiments, unstretched myocyte cultures were incubated for 1 hour in serum-free medium containing exogenous recombinant human VEGF165 (the 165 amino acid isoform of human VEGF; 100 ng/mL, R&D Systems), a cyclic RGD peptide (GRDGSP) or an inactive RGE peptide (GRGESP) (0.5 mmol/L each, Washington University Protein Chemistry Laboratory), or MnCl2 (1 mmol/L, Sigma). In other experiments, cultures were pretreated for 30 minutes in serum-free media containing 50 e/mL monoclonal anti-1 (clone Ha2/5, BD Pharmingen) or anti-3 integrin antibodies (clone 2C9.G2, BD Pharmingen) before being subjected to pulsatile stretch or incubation with other reagents. These antibodies have been shown in previous studies to block signaling mediated by 1 and 3 integrins, respectively.10,11 Control cultures (not stretched or exposed to exogenous reagents) were also incubated in serum-free medium for 1 hour. At the conclusion of each experimental protocol, cells were either scraped from membranes and lysed for immunoblot analysis or fixed for confocal immunofluorescence microscopy.
Immunoblot Assay of Cx43 Expression
Cell cultures were washed once in cold PBS and then scraped from silicone membranes in a low ionic strength buffer containing protease inhibitors (NaHCO3,1 mmol/L; EDTA, 5 mmol/L; EGTA, 1 mmol/L; leupeptin, 1 eol/L; pepstatin, 1 eol/L; aprotinin, 0.1 eol/L; benzamidine, 1 mmol/L; iodoacetamide, 1 mmol/L; phenylmethylsulfonylfluoride, 1 mmol/L). After centrifugation at 3000g for 6 minutes at 4°C, the pellet was resuspended in the same buffer and stored at eC70°C. Aliquots containing 7 e of total protein were analyzed by SDS polyacrylamide gel electrophoresis, and Cx43 signal was quantified by densitometry as described in detail in previous studies.12
Quantitative Confocal Immunofluorescence Microscopy
Cultures were rinsed in serum-free medium and then fixed in 4% paraformaldehyde. To measure the amount of Cx43 immunoreactive signal located in gap junctions, cells were incubated overnight with a polyclonal rabbit anti-Cx43 antibody (diluted 1:400; product no. 71-0700, Zymed) and then with Cy3-conjugated goat anti-rabbit IgG (diluted 1:400, Jackson Immunoresearch) before being examined by laser scanning confocal microscopy. In selected experiments, expression of N-cadherin was analyzed using a rabbit polyclonal antibody that recognizes a common epitope in the N-cadherins (product no. C3678, Sigma). The area occupied by specific Cx43 or N-cadherin immunoreactive signal in cell-cell junctions was quantified by digital image processing and expressed as a percentage of total myocyte area using established, validated methods that have been described previously in detail.13 In each culture, three fields were randomly selected for quantitative analysis. Each field was 26 400 e2 in area and included profiles of 30 to 50 cells. A mean value was calculated for each culture by averaging the results from the three fields. At least three individual cultures were analyzed for each condition (n3 experiments). The amount and distribution of 1 integrin signal was also characterized by immunofluorescence microscopy in paraformaldehyde-fixed cells, which were incubated overnight with an Armenian hamster anti-1 integrin antibody (diluted 1:1000, BD Pharmingen) and then with Cy3-conjugated goat anti-Armenian hamster IgG (diluted 1:400, Jackson Immunoresearch).
Dye-Transfer Assay
Cells grown on silicone membranes coated with type I collagen or fibronectin (n=3 membranes for each) were rinsed in PBS and then incubated with Lucifer yellow (1%, Sigma). A small glass-cutting wheel was used to make a clean cut across the myocyte monolayer to permit the membrane-impermeant dye to enter a linear array of broken cells (scrape-loading of dye). After being incubated for 2 minutes at 37°C, the preparations were rinsed three times with PBS, fixed in 4% paraformaldehyde, mounted on glass slides, and photographed using a fluorescence microscope and Nikon Coolpix 995 camera. Digital images were analyzed off-line using ImageJ software (NIH, http://rsb.info.nih.gov/ij/). Three fields from each culture were analyzed. An area of dye transfer adjacent to the scrape line was obtained from each field and divided by the length over which the area was measured, to yield an average linear value of dye-spread expressed in micrometers (e).
Statistical Analysis
Data are expressed as mean±SD. Differences between groups were analyzed with ANOVA and Fisher protected least significant difference test. A value of P<0.05 was considered sufficient to reject the null hypothesis.
Results
Myocyte-Matrix Protein Interactions Affect Cx43 Expression at Gap Junctions
To determine whether the chemical composition of the extracellular matrix influences Cx43 expression in cardiac myocytes grown under basal conditions (without stretch), cultures were plated on either native type I collagen or fibronectin, grown for 4 days, and Cx43 expression at cell-cell junctions was measured by quantitative confocal microscopy. As shown in representative confocal immunofluorescent images and quantitative confocal analysis of three independent sets of cultures (Figure 1), the amount of Cx43 immunoreactive signal at sites of intercellular junctions was 2-fold greater in myocyte cultures grown on fibronectin than on collagen (2.19%±0.61 versus 1.05%±0.13, respectively; P<0.003). Additional cells grown on native collagen or fibronectin were subjected to pulsatile stretch or incubated with exogenous VEGF for 1 hour. As previously shown,3,4 myocytes plated on collagen showed a significant increase in the amount of Cx43 immunoreactive signal at intercellular junctions in response to pulsatile stretch (from 1.05±0.13% to 2.08±0.74%; P<0.04) and after being incubated with VEGF for 1 hour (from 1.05±0.13% to 2.48±0.60%; P<0.02) (Figure 1). However, neither stretch nor VEGF increased the amount of Cx43 immunoreactive signal in myocytes plated on fibronectin (Figure 1). The amount of Cx43 expression in cells grown on fibronectin was roughly equal to the level observed after stretch or incubation with VEGF in cells grown on native type I collagen. No apparent change in cell shape was noted in cells grown on fibronectin or subjected to stretch.
Previously, we showed that enhanced Cx43 expression in response to stretch in cells grown on collagen was associated with increased coupling.3,4 To determine whether increased Cx43 expression in nonstretched cells grown on fibronectin was also associated with enhanced coupling, we scrape-loaded cells with the membrane-impermeant dye, Lucifer yellow, and measured the extent to which dye passed to adjacent cells through gap junctions. As shown in Figure 1, dye traveled 71% further in nonstretched cells grown on fibronectin (229±35 e) than in nonstretched cells grown on collagen (134±41 e; n=3 for each, P<0.04).
Integrin-Matrix Protein Interactions Alter Cx43 Expression at Gap Junctions
Fibronectin contains abundant RGD cell adhesive motifs.13,14 Furthermore, myocardial injury during ischemia or infarction leads to exposure of previously cryptic RGD motifs in native collagen in the myocardial interstitium.8,9,14eC16 To determine whether denaturation of collagen, which is known to expose RGD motifs, would produce changes in Cx43 expression similar to those seen in cells grown on fibronectin, myocytes were grown on either native collagen or collagen subjected to heat-denaturation. As shown in representative confocal immunofluorescent images and quantitative confocal analysis from three independent experiments (Figure 2), the amount of Cx43 immunoreactive signal at sites of intercellular junctions was 1.6-fold greater in myocytes grown on heat-denatured collagen under basal conditions than in myocytes grown under identical conditions on native collagen (1.96%±0.63 versus 1.19%±0.28, respectively; P<0.03). As was observed in cells grown on fibronectin, myocytes plated on heat-denatured collagen showed no significant increase in Cx43 signal in response to 1 hour of pulsatile stretch or incubation with exogenous VEGF (Figure 2). Immunoblotting demonstrated an increase in the total amount of Cx43 in response to stretch in myocytes grown on native collagen (Figure 2), indicating that enhanced Cx43 signal at cell-cell junctions seen by confocal microscopy reflected an increase in the total Cx43 content within cardiac myocytes rather than only translocation from intracellular to junctional pools. A similarly increased level of Cx43 expression was also seen in nonstretched cells grown on denatured collagen, and there was no further increase after stretch (Figure 2). No apparent changes in Cx43 phosphorylation were observed, at least as judged by the relative intensities of the various Cx43 bands on immunoblots. No changes in cell shape were noted either.
Previously, we found that expression of the intercellular adhesion molecule N-cadherin was increased in response to stretch in cells grown on native type I collagen.3 To determine whether expression of this mechanical junction protein was also affected by different matrix proteins, cells grown on fibronectin and denatured collagen were subjected to stretch and N-cadherin expression was measured by confocal microscopy. As shown in Figure 3, N-cadherin expression was increased in nonstretched cells grown on fibronectin or denatured collagen compared with nonstretched cells grown on native collagen. Stretch increased N-cadherin expression by >3-fold in cells on native collagen but had no effect in cells grown on fibronectin or denatured collagen. These results suggest that expression of electrical and mechanical junction proteins is coordinately regulated.
Results in cells grown on fibronectin or denatured collagen suggest that integrin signaling initiated by RGD motifs upregulates Cx43 expression and increases the amount of Cx43 in gap junctions independent of stretch and VEGF. To further determine the effect of the RGD motif on Cx43 expression, myocytes plated on native type I collagen were incubated with a cyclic RGD peptide (GRGDSP). Although this peptide can be used to block integrin signaling, we observed a marked increase in Cx43 expression. As shown in representative confocal images and by quantitative confocal analysis (Figure 4), the amount of Cx43 immunoreactive signal at intercellular junctions was >2-fold greater (2.57±0.60% versus 1.10±0.06%, respectively; P<0.002) in myocytes incubated with the RGD peptide than in control cultures not treated with peptide. Incubation of cells with an RGE peptide had no effect (data not shown). Because cellular expression and recruitment of specific integrin heterodimers occurs with RGD exposure, these findings suggest that activation of integrin heterodimers alters myocyte expression of Cx43.
Generalized Activation of Integrins Increases Cx43 Expression at Gap Junctions
To further investigate the role of integrin signaling in regulating Cx43 expression, we exposed cells to MnCl2. The divalent cation Mn2+ stabilizes an open, extended integrin conformation promoting constitutive, high-affinity ligand binding.17 To determine whether generalized integrin activation affects Cx43 expression at gap junctions, myocytes grown on native type I collagen or heat-denatured collagen were exposed to 1 mmol/L MnCl2 for 1 hour. No change in pH was detected in the media of cells incubated with MnCl2. The amount of Cx43 immunoreactive signal at intercellular junctions was significantly increased in myocytes on native collagen that were incubated with MnCl2 (from 1.28±0.14% to 2.11±0.20%; P<0.001) but not in cells grown on heat-denatured collagen (Figure 5). Immunoblotting demonstrated an increase in the total content of Cx43 in myocytes incubated with MnCl2 (Figure 5) Thus, generalized activation of integrins upregulates Cx43 expression and increases the amount of Cx43 in gap junctions independent of mechanical perturbation.
Role for 1 Integrin Signaling in Mechanoelectrical Coupling
1 Integrins have been shown to play an essential role as mechanotransducers in the heart.1,2,18eC21 Although cardiac myocytes express both 1 and 3 integrin subunits,1,2 native collagen binds only to integrin heterodimers containing the 1 integrin subunit.22 To further elucidate the respective roles of 1 and 3 integrins in regulating Cx43 expression, myocytes grown on native collagen were exposed to MnCl2 in the presence or absence of monoclonal antibodies against 1 or 3 integrin subunits. As previously demonstrated, addition of MnCl2 significantly increased Cx43 expression at gap junctions (2.21±0.42% versus 1.05±0.21%, respectively; P<0.04) compared with controls (Figure 6). Pretreatment with anti-1 integrin antibody completely blocked the increase in Cx43 expression at gap junctions caused by MnCl2 (1.04±0.04% versus 1.05±0.21%, respectively; P=NS) (Figure 6). In contrast, anti-3 integrin antibody had no effect on the degree of upregulation of Cx43 expression induced by MnCl2 (Figure 6). Immunoblotting showed that pretreatment with anti-1 integrin antibody apparently decreased total Cx43 content in cells incubated with MnCl2 despite the fact that Cx43 signal in junctions was not different than in control cells (Figure 6). The significance of this specific observation is unclear but overall, the results of this experiment indicate that generalized activation of integrin signaling by MnCl2 upregulates Cx43 expression by a 1 integrin-specific mechanism.
To further characterize the role of 1 integrins in mechanotransduction, myocytes grown on native collagen were subjected to pulsatile stretch for 1 hour and the amount and distribution of 1 integrin immunofluorescent signal was analyzed by confocal microscopy. As shown in Figure 7, the amount of 1 immunoreactive signal was markedly increased in response to stretch. However, no apparent increase in 1 integrin signal was noted in unstretched cells grown on fibronectin or denatured collagen, or in cells on native collagen that were incubated with Mn2+. Finally, to determine whether 1 integrin signaling mediates stretch-induced upregulation of Cx43 expression, cells grown on native collagen were stretched in the presence or absence of anti-1 integrin antibody. As previously shown, stretch led to significant upregulation of Cx43 expression (from 1.05±0.07% to 1.91±0.18%, n=4; P<0.001) (Figure 8). However, pretreatment with anti-1 integrin antibody blocked stretch-induced upregulation of Cx43 expression (Figure 8). Addition of exogenous VEGF to nonstretched cells that had been pretreated with anti-1 integrin antibody upregulated Cx43 expression (to 1.83%±0.40, n=4; P<0.02 versus control cells) (Figure 8). Thus, blockade of integrin signaling does not prevent cells from responding to VEGF.
Discussion
The present study was designed to elucidate the effects of specific matrix proteins in regulating cardiac myocyte gene expression induced by mechanical load. We compared the effects of short-term pulsatile stretch on ventricular myocytes grown on native type I collagen, the major component of the extracellular matrix in the normal heart, and myocytes grown on fibronectin and denatured collagen, both of which accumulate in the myocardial interstitium under selected pathophysiological conditions. We focused specifically on effects of mechanical load on expression of Cx43, the major cardiac gap junction channel protein, in ventricular myocytes grown on the different matrix protein. In previous studies,3 we observed that Cx43 expression increased by 2-fold after 1 hour of pulsatile stretch in ventricular myocytes grown on type I collagen. We also showed previously that VEGF, secreted during stretch, acted through an autocrine mechanism to upregulate Cx43 expression, and that the stretch effect could be mimicked by incubating nonstretched cells with exogenous VEGF.4 In the present study, we found that when ventricular myocytes were grown on either fibronectin or denatured collagen, they expressed increased amounts of Cx43 compared with cells grown on type I collagen. Cx43 levels in nonstretched cells on fibronectin or denatured collagen were roughly equivalent to the levels seen after stretch in cells grown on native type I collagen. Moreover, neither stretch nor exogenous VEGF increased Cx43 expression in cells grown on fibronectin or denatured collagen. Similar results were observed for the intercellular adhesion molecule, N-cadherin. These observations suggest that signaling pathways activated by stretch or VEGF in cells grown on native collagen are constitutively activated in cells grown on fibronectin or denatured collagen such that no further increase in Cx43 expression is caused by stretch or VEGF.
In selected experiments, we observed that increased Cx43 expression at cell-cell junctions was accompanied by an increase in total Cx43 content measured by immunoblotting. This confirms previous findings.4 We have also shown previously that increased Cx43 expression induced by stretch3,4 or by exposure to cAMP23 in vitro leads to enhanced coupling as demonstrated by increased conduction velocity. In the present study, we showed that increased Cx43 expression in cells grown on fibronectin was associated with enhanced junctional coupling as demonstrated using a standard dye-transfer assay. These results provide further evidence that upregulation of Cx43 expression at cell-cell junctions, measured by confocal microscopy, reflects an increase in functional gap junction channels.
Because both fibronectin and denatured collagen, but not native collagen, interact with integrins via RGD motifs,8,9,14eC16 we reasoned that activation of integrin signaling by RGD was responsible for enhanced Cx43 expression under basal conditions in cells grown on fibronectin or denatured collagen. Indeed, we could mimic the stretch effect by incubating cells on native collagen with a cyclic RGD peptide to specifically activate integrins dependent on RGD signaling, or by activating integrins nonspecifically with MnCl2. No further increase in Cx43 expression was achieved by exposing cells on fibronectin or denatured collagen to MnCl2. Thus, integrin signaling responsible for upregulating Cx43 expression appears to be maximal in cells grown on fibronectin or denatured collagen.
Finally, we showed that enhanced Cx43 expression induced by incubating cells grown on native collagen with MnCl2 could be blocked completely by an anti-1 integrin antibody but not an anti-3 integrin antibody. Moreover, stretch-induced upregulation of Cx43 expression was also blocked by anti-1 integrin antibody in cells grown on native collagen. These results indicate that signaling activated by mechanical load in cardiac myocytes is mediated by 1 integrins. The fact that the effects of Mn2+ was blocked by anti-1 integrin antibody also suggests that Mn2+ increased Cx43 expression by activating integrin signaling rather than by other potential mechanisms.
Several classes of molecules may function as mechanosensors to regulate cardiac myocyte responses to mechanical load. Previous studies have identified only a limited role of stretch-activated ion channels (pacemaker inward, ATP-sensitive K+, and stress sensitive cation channels) in sensing mechanical stimuli and transducing a hypertrophic response in myocytes.24,25 Whereas intercellular adhesion molecules (ICAMs) also activate signaling pathways,26 only the integrins directly connect the extracellular matrix to the actin cytoskeleton, and many studies have demonstrated that transduction of mechanical stress in myocytes into biochemical signals is mediated by integrins.1,2
Type I and III collagens constitute the great majority of the extracellular matrix in the normal heart, but production and secretion of fibronectin by fibroblasts is a key early step in establishment of the matrix in the developing heart.27,28 Fibronectin interacts with collagen, heparin, fibrin, and cell surface integrins to bind diverse matrix components and serve as a scaffold for nascent matrix assembly.15 In the present studies, nearly pure populations of cardiac myocytes were grown on defined matrix proteins to delineate specific matrix-myocyte interactions. Because contamination by fibroblasts is low (<5%),4 the effects of fibronectin and/or other matrix components secreted by fibroblasts in cultures plated on collagen were probably negligible. Similarly, there were no apparent effects of fibronectin that may have been derived from serum in cell culture media.
Various pathological conditions are associated with progressive qualitative and quantitative changes in the composition of the extracellular matrix, which may lead to interstitial fibrosis.27,29,30 Increased levels of mRNAs encoding fibronectin and collagen have been observed during the transition from hypertrophy to failure in spontaneously hypertensive rats and in pressure-overload hypertrophy induced by aortic constriction in rats.5eC7 Similarly, denaturation and matrix metalloproteinaseeCmediated degradation of collagen induced by acute myocardial ischemia have been shown to expose previously cryptic RGD motifs within interstitial type I collagen, and this may be followed by increased expression of fibronectin.8,9 Although the defined in vitro conditions used in the present studies cannot precisely reflect changes in the matrix in cardiac disease, our observations provide new insights about how matrix remodeling and RGD-mediated activation of integrins might activate signaling cascades that promote adaptive responses to injury.
Enhanced Cx43 expression is known to occur during early, compensatory phases of cardiac hypertrophy,31 but the potential role of altered myocyte-matrix interactions in this context are unknown.32 Remodeling of gap junctions is also well documented in regions bordering acute myocardial infarcts. The appearance of RGD sites within discrete areas of acute injury may help regulate the repair process and/or initiate adaptive responses in viable but injured cells adjacent to the zone of necrosis. Alternatively, nonuniform, heterogeneous remodeling of the extracellular matrix might contribute to development of discrete areas of differential Cx43 expression and account for enhanced anisotropic conduction and formation of functional lines of block, which could promote reentrant arrhythmias.33 The present study provides evidence for a potential new mechanism linking remodeling of the extracellular matrix to changes in electrical coupling of cardiac myocytes.
The 1 integrin is the predominant -subunit found in the myocardium and has been implicated in mediating the cardiac hypertrophic response.18eC22 A crucial role for 1 integrin signaling in sensing shear stress in endothelial cells has also been well established.34 Our findings that 1 integrin expression increases with pulsatile stretch (a result that has been demonstrated previously with static stretch)35 and that 1 integrin signaling is responsible for upregulation of Cx43 induced by MnCl2 provide additional evidence for a primary role of 1 integrins as mechanotransducers in cardiac myocytes. Our observations suggest further that altered expression of Cx43 and remodeling of intercellular electrical junctions could play a role in the pathogenesis of the lethal phenotype observed in 1D integrin knockout mice after transaortic constriction.20
In conclusion, we have shown that matrix protein-myocyte interactions can regulate Cx43 expression in gap junctions via 1 integrin signaling initiated by mechanical stimulation in cells grown on native type I collagen, or by RGD-integrin signaling independent of mechanical stress in cells grown on fibronectin or denatured collagen. The composition of the extracellular matrix in physiological or pathophysiological states may, therefore, affect the degree of electrical coupling in cardiac myocytes.
Acknowledgments
This work was supported by NIH grants HL50598 and HL74595. We thank Dr Steven Teitelbaum for helpful discussions and critical review of the manuscript.
References
Katsumi A, Orr AW, Tzima E, Schwartz MA. Integrins in mechanotransduction. J Biol Chem. 2004; 279: 12001eC12004.
Ross RS, Borg TK. Integrins and the myocardium. Circ Res. 2001; 88: 1112eC1119.
Zhuang J, Yamada KA, Saffitz JE, Kleeber AG. Pulsatile stretch remodels cell-to-cell communication in cultured myocytes. Circ Res. 2000; 87: 316eC322.
Pimentel RC, Yamada KA, Kleeber AG, Saffitz JE. Autocrine regulation of myocyte Cx43 expression by VEGF. Circ Res. 2002; 90: 671eC677.
Villareal FJ, Dillman WH. Cardiac hypertrophy-induced changes in mRNA levels for TGF-1, fibronectin, and collagen. Am J Physiol. 1992; 31: H1861eCH1866.
Boluyt MO, O’Neill L, Meredith AL, Bing OH, Brooks WW, Conrad CH, Crow MT, Lakatta EG. Alterations in cardiac gene expression during the transition from stable hypertrophy to heart failure: marked upregulation of genes encoding extracellular matrix proteins. Circ Res. 1994; 75: 23eC32.
Boluyt MO, Bing OH. Matrix gene expression and decompensated heart failure: the aged SHR model. Cardiovasc Res. 2000; 46: 239eC249.
Cannon RO III, Butany JW, McManus BM, Speir E, Kravitz AB, Bolli R, Ferrans VJ. Early degradation of collagen after acute myocardial infarction in the rat. Am J Cardiol. 1983; 52: 390eC395.
Davis GE, Bayless KJ, Davis MJ, Meininger GA. Regulation of tissue injury responses by the exposure of matricryptic sites within extracellular matrix molecules. Am J Pathol. 2000; 156: 1489eC1498.
Schultz JF, Armant DR. Beta 1- and beta 3-class integrins mediate fibronectin binding activity at the surface of developing mouse peri-implantation blastocysts: regulation by ligand-induced mobilization of stored receptor. J Biol Chem. 1995; 270: 11522eC11531.
Ogawa E, Saito Y, Harada M, Kamitani S, Kuwahara K, Miyamoto Y, Ishikawa M, Hamanaka I, Kajiyama N, Takahashi N, Nakagawa O, Masuda I, Kishimoto I, Nakao K. Outside-in signaling of fibronectin stimulates cardiomyocyte hypertrophy in cultured neonatal rat ventricular myocytes. J Mol Cell Cardiol. 2000; 32: 765eC776.
Johnson CM, Green KG, Kanter EM, Bou-Abboud E, Saffitz JE, Yamada KA. Voltage-gated Na2+ channel activity and connexin expression in Cx43-deficient myocytes. J Cardiovasc Electrophysiol. 1999; 10: 1390eC1401.
Saffitz JE, Green KG, Kraft WJ, Schechtman KB, Yamada KA. Effects of diminished expression of connexin43 on gap junction number and size in ventricular myocardium. Am J Physiol Heart Circ Physiol. 2000; 278: H1662eCH1670.
Yamamoto K, Yamamoto M. Cell adhesion receptors for native and denatured type I collagens and fibronectin in rabbit arterial smooth muscle cells in culture. Exp Cell Res. 1994; 214: 258eC263.
Wierzbicka-Patynowski I, Schwarzbauer JE. The ins and outs of fibronectin matrix assembly. J Cell Sci. 2003; 116: 3269eC3276.
Davis GE. Affinity of integrins for damaged extracellular matrix: alpha v beta 3 binds to denatured collagen type I through RGD sites. Biochem Biophys Res Commun. 1992; 182: 1025eC1031.
Takagi J, Petre BM, Walz T, Springer TA. Global conformational rearrangements I integrin extracellular domains in outside-in and inside-out signaling. Cell. 2002; 110: 599eC611.
Ross RS, Pham C, Shai SY, Goldhaber JI, Fenczik C, Glembotski CC, Ginsberg MH, Loftus JC. Beta1 integrins participate in the hypertrophic response of rat ventricular myocytes. Circ Res. 1998; 82: 1160eC1172.
Pham CG, Harpf AE, Keller RS, Vu HT, Shai SY, Loftus JC, Ross RS. Striated muscle-specific beta(1D)-integrin and FAK are involved in cardiac myocyte hypertrophic response pathway. Am J Physiol Heart Circ Physiol. 2000; 279: H2916eCH2926.
Shai SY, Harpf AE, Babbitt CJ, Jordan MC, Fishbein MC, Chen J, Omura M, Leil TA, Becker KD, Jiang M, Smith DJ, Cherry SR, Loftus JC, Ross RS. Cardiac myocyte-specific excision of the beta1 integrin gene results in myocardial fibrosis and cardiac failure. Circ Res. 2002; 90: 458eC464.
Brancaccio M, Fratta L, Notte A, Hirsch E, Poulet R, Guazzone S, De Acetis M, Vecchione C, Marino G, Altruda F, Silengo L, Tarone G, Lembo G. Melusin, a muscle-specific integrin beta1-interacting protein, is required to prevent cardiac failure in response to chronic pressure overload. Nat Med. 2003; 9: 68eC75.
Gullberg DE, Lundgren-Akerlund E. Collagen-binding I domain integrins: what do they do Prog Histochem Cytochem. 2002; 37: 3eC54.
Darrow BJ, Fast VG, Kleeber AG, Beyer EC, Saffitz JE. Functional and structural assessment of intercellular communication: increased conduction velocity and enhanced connexin expression in dibutyryl cAMP-treated cultured cardiac myocytes. Circ Res. 1996; 79: 174eC183.
Sadoshima J, Takahashi T, Jahn L, Izumo S. Roles of mechano-sensitive ion channels, cytoskeleton, and contractile activity in stretch-induced immediate-early gene expression and hypertrophy of cardiac myocytes. Proc Natl Acad Sci U S A. 1992; 89: 9905eC9909.
Yamazaki T, Komuro I, Kudoh S, Zou Y, Nagai R, Aikawa R, Uozumi H, Yazaki Y. Role of ion channels and exchangers in mechanical stretch-induced cardiomyocyte hypertrophy. Circ Res. 1998; 82: 430eC437.
Thiery JP. Cell adhesion in development: a complex signaling network. Curr Opin Genet Dev. 2003; 13: 365eC371.
Weber KT, Brilla CG. Pathological hypertrophy and the cardiac interstitium. Circulation. 1991; 83: 1849eC1865.
Raghow R. The role of extracellular matrix in postinflammatory wound healing and fibrosis. FASEB J. 1994; 8: 823eC831.
Lindsey ML, Mann DL, Entman ML, Spinale FG. Extracellular matrix remodeling following myocardial injury. Ann Med. 2003; 35: 316eC326.
Janicki JS, Brower GL, Gardner JD, Chancey AL, Stewart JA Jr. The dynamic interaction between matrix metalloproteinase activity and adverse myocardial remodeling. Heart Fail Rev. 2004; 9: 33eC42.
Kanno S, Saffitz JE. The role of gap junctions in electrical conduction and arrhythmogenesis in the heart. Cardiovasc Pathol. 2001; 10: 169eC177.
Peters NS, Coromilas J, Severs NJ, Wit AL. Disturbed connexin43 gap junction distribution correlates with the location of reentrant circuits in the epicardial border zone of healing canine infarcts that cause ventricular tachycardia. Circulation. 1997; 95: 988eC996.
van Rijen HV, Eckardt D, Degen J, Theis M, Ott T, Willecke K, Jongsma HJ, Opthof T, de Bakker JM. Slow conduction and enhanced anisotropy increase the propensity for ventricular tachyarrhythmias in adult mice with induced deletion of connexin43. Circulation. 2004; 109: 1048eC1055.
Shyy JY, Chien S. Role of integrins in endothelial mechanosensing of shear stress. Circ Res. 2002; 91: 769eC775.
Sharp WW, Simpson DC, Borg TK, Samarel AM, Terracio L. Mechanical forces regulate focal adhesion and costamere assembly in cardiac myocytes. Am J Physiol Heart Circ Physiol. 1997; 273: H546eCH556.(Amit J. Shanker, Kiyomi Y)