- Email: [email protected]
Cardiac Integrins
SYMPOSIUM
Cardiac Integrins: The Ties That Bind David G. Simpson, Titus A. Reaves, Daw-tsun Shih, William Burgess, Thomas K. Borg, and Louis Terracio Department of Developmental Biology and Anatomy, School of Medicine, University of South Carolina, Columbia, South Carolina
11 An elaborate series of morphogenetic events must be precisely coordinated during development to promote the formation of the elaborate three-dimensional structure of the normal heart. In this study we focus on discussing how interconnections between the cardiac myocyte and its surrounding environment regulate cardiac form and function. In vitro experiments from our laboratories provide direct evidence that cardiac cell shape is regulated by a dynamic interaction between constituents of the extracellular matrix (ECM) and by specific members of the integrin family of matrix receptors. Our data indicates that phenotypic information is stored in the tertiary structure and chemical identity of the ECM. This information appears to be actively communicated and transduced by the a1b1 integrin molecule into an intracellular signal that regulates cardiac cell shape and myofibrillar organization. In this study we have assessed the phenotypic consequences of suppressing the expression and accumulation of the a1 integrin molecule in aligned cultures of cardiac myocytes. In related experiments we have examined how the overexpression of a2 and a5 integrin, integrins normally not present or present at very low copy number on the cell surface of neonatal cardiac myocytes, affect cardiac protein metabolism. We also consider how biochemical signals and the mechanical signals mediated by the integrins may converge on common intracellular signaling pathways in the heart. Experiments with the whole embryo culture system indicate that angiotensin II, a peptide that carries information concerning cardiac load, plays a role in controling cardiac looping and the proliferation of myofibrils during development. Cardiovasc Pathol 1998;7:135–143 © 1998 by Elsevier Science Inc.
The formation of the heart involves a series of well orchestrated morphogenetic events that must occur in a precisely coordinated sequential order. The developmental processes of commitment, migration, differentiation, and morphogenesis must all be integrated during fetal life to produce the elaborate three-dimensional structure of the heart (1,2). Deviations in this developmental program result in a variety of congenital malformations. These malformations range from relatively benign abnormalities in ventricular architecture to profound defects that result in the death of the developing fetus (3,4). The initial steps of early differentiation have come under intense investigation. However, a complete unManuscript received/accepted 20 November 1997. Presented as part of the International Society of Heart Research meeting, Vancouver, 1997 Address for correspondence: Louis Terracio, Department of Developmental Biology and Anatomy, University of South Carolina School of Medicine, Columbia, SC 29208; phone (803) 733-3369. Cardiovascular Pathology Vol. 7, No. 3, May/June 1998:135–143 1998 by Elsevier Science Inc. All rights reserved. 655 Avenue of the Americas, New York, NY 10010
derstanding of the regulatory processes that direct cardiac morphogenesis and the expression of cardiac cell phenotype has not been achieved. We focus this review on discussing how interconnections between the cell and its surrounding environment regulate cardiac form and function. Typically, the interconnections between the cell and its surrounding environment are regarded as a physical phenomenon, however, in many cases physical perturbations and the biochemical signals that link the heart to the surrounding interstitium and the other organ systems converge at a common intracellular second messenger pathway. We discuss some of potential sites where these signals converge. Fundamental principals of developmental biology, derived from elegant experimental paradigms in flies, nematodes, and fish are being applied to mammalian systems (5,6). Classical experiments demonstrated that developmental signals must be transmitted and received in a precise sequential fashion. As pointed out by Raff (6) the current
1054-8807/98/$19.00 PII S1054-8807(97)00123-3
136
SIMPSON ET AL. CARDIAC INTEGRINS
challenge is to understand how extracellular signals and intracellular events interact to regulate the differentiation of individual cells and the maturation of organ systems. The structural and functional profile of a cardiac myocyte undergoes a profound change during the developmental process. In the earliest stages of embryonic life the presumptive myocyte must migrate a considerable distance to assume its proper position within the condensing cell mass that will eventually form the heart. During this stage of development the presumptive cardiac myocyte has a nondescript cell shape, the capacity to proliferate, and lacks myofibrils. As development proceeds, the myocyte differentiates, assembles myofibrils, loses its capacity to divide, and assumes a rod-like cell shape. In order to fully understand how the distinctive phenotype of the adult cardiac myocyte is established and maintained, we must define the spatial and temporal sequence of events that direct cardiac development. The term phenotype is usually used to define the structural appearance, or shape, of a given cell type. However, in this discussion we use this term to describe both the structural and biochemical features that determine cell shape and function. We have adopted this convention because the structural characteristics of a cell are ultimately defined by its genetic potential and the biochemical processes that occur within the intracellular environment. From this definition it is clear that the evolution of a particular phenotype is the result of a dynamic interaction between the intrinsic genetic program of a cell and factors in the surrounding environment that regulate the extent to which this genetic program is implemented. Specific gene sequences can be regulated by the composition and arrangement of the surrounding extracellular matrix (ECM), mechanical signals, cell-to-cell communication, and a variety of soluble neurohumoral factors. It has become increasingly evident that a variety of cellular functions are intimately linked to the phenotype that a given cell type expresses (7,8). For example, in vitro many polarized secretory cells will not synthesize the appropriate complement of cellular proteins unless they are induced to “differentiate” and develop a basal, lateral, and apical surface (9,10). When fibroblasts are cultured in a three-dimensional matrix, instead of a planer surface, they express a more fusiform cell shape, display reduced proliferative capacity (11– 13), and alter their rate of collagen synthesis to reflect a more in vivo-like pattern (14). In the converse situation, if the normal structural phenotype of a cell is altered by an outside influence its ability to function may be drastically impaired. The scarring that occurs in association with an infarct compromises the structural integrity of the ventricular wall, modifies the conductive properties of the tissue, and alters the ability of the surrounding muscle cells to effectively contract.
Signals Associated with ECM In the normal adult myocardium the ECM is organized into a complex, three-dimensional network that envelops
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143
the individual cells of the heart (15–18). Selected components of this network are anchored to the surface of the myocytes through specific binding reactions (Figure 1). The ECM is produced largely by the fibroblasts, and its composition and arrangement is subject to change during development and normal aging. The biochemical and structural complexity of the ECM evolves gradually and in parallel with the developing heart (19–21). The chemical identity of the ECM also can be altered in disease states. Notably, during the onset of a pressure overload the ratio of type I to type III collagen is altered and the heart becomes less compliant (22–25). This has obvious consequences and alters cardiac contractile function on a global scale. Signals communicated from the ECM to the cell surface have been implicated in the regulation of biological events associated with cardiac development and the maintenance of the normal adult phenotype (26,27). In vitro experiments from our laboratories provide direct evidence that cardiac cell shape can be regulated by the composition and arrangement of the surrounding extracellular matrix. We have developed a substrate for cell culture composed of a thin film of aligned collagen fibrils (27). This matrix is prepared by applying a solution of ice cold, neutral collagen to a culture dish along a single, unique axis. The collagen solution is then drained from the culture substratum along the axis in which it was applied and the dish is then placed into a 378C incubator to polymerize the col-
Figure 1. Schematic of the ECM-integrin-cytoskeletal connection depicting the physical connection of the cytoskeleton to the ECM through the integrin complex. This continuum allows for the transmission of contractile forces from the myofibrils out into the ECM. Conversely, physical perturbations and other phenotypic signals can be communicated from the ECM to the intracellular space via the ECM-integrin-cytoskeletal connection. These external signals may be transmitted directly by mechanical ques or, indirectly, by biochemical signal cascades initiated by the integrin and/or molecules associated with the cytoplasmic tail of the integrin.
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143
lagen. The resulting matrix consists of collagen fibrils aligned along a common axis. When fetal (28) or neonatal cardiac myocytes (27) are cultured on this type of substrate they spread, as a population, in parallel with the fibrils of the underlying matrix. The individual cells of these cultures have an elongated, rod-like cell shape with myofibrils that are distributed in parallel with their long axis (Figure 2). In contrast, when neonatal cardiac myocytes are cultured on plastic, laminin, fibronectin, or a random film of collagen they express a flattened, stellate cell shape (27). Our experiments have demonstrated that a critical concentration of collagen is necessary to promote the expression of the rodlike phenotype. Applying too little, or too much, collagen to the dish disrupts the formation of the elongated cell shape and the tissue-like pattern of organization (27). Altering the profile of carbohydrates present on the collagen fibrils (unpublished observation, TA Reeves) or overcoating the collagen with fibronectin or laminin will also disrupt the expression of the aligned phenotype (27).
Integrin-ECM Connection The phenotypic information stored in the tertiary structure and chemical identity of the ECM is actively communicated to the cardiac myocyte. Constituents of the ECM are physically attached to the surface of these cells through specific interactions mediated by the integrin family of matrix receptors (29). The integrins are transmembrane molecules that are composed of an a and a b subunit. Various a and b chains combine to form receptors for different constituents of the ECM. Within the intracellular space, the cytoplasmic domains of the integrin complex interact with elements of the cytoskeleton (30), the internal scaffolding that supports
SIMPSON ET AL. CARDIAC INTEGRINS
137
and surrounds the myofibrils (Figure 1). Physical perturbations to the integrin complex are associated with the activation of phosphotyrosine cascades (30–32). Thus, external signals may be physically or biochemically communicated to the intracellular space by the integrin complex.
Integrin-ECM Signaling It is postulated that integrins transduce phenotypic information to cardiac myocytes in response to developmental events and disease processes that alter the composition or arrangement of the ECM. Integrins of the b1 subfamily are expressed in precise temporal and spatial patterns during embryogenesis in several organ systems including the integument (33) and kidney (34). This subfamily is also the predominate isoform of the myocardium (29). In embryonic life b1 mRNA and protein are expressed most abundantly in the developing trabeculae and cardiac cushions (35). As development proceeds, b1 protein accumulates in the ventricular myocytes and the leaflets of the valves. Preliminary evidence also indicates that the a chains of the integrin complex are expressed in very specific patterns in the developing heart (36). For instance, the integrin receptors for laminin and fibronectin are sequentially and regionally expressed in early heart development (37). With time this integrin subunit accumulates in the myocytes and subsequently persists in both the muscle and endothelial cells throughout embryonic and fetal life. The a7 chain of the integrin complex does not appear until late in fetal development and is restricted almost exclusively to the developing myocytes of the ventricular wall (38). It has been difficult to determine the functional role of individual integrin molecules during cardiac development.
Figure 2. Micrographs of the aligned myocyte phenotype. Cells exhibit an elongated cell shape and are distributed as a population along a common axis in a tissue-like pattern of organization (A). The underlying collagen fibrils (not visible) are arrayed across the long axis of the micrograph from right to left. The myofibrils of these rod-like cells are distributed in parallel with one another (B). Myocytes plated onto a non-aligned matrix of collagen express a stellate cell shape and random arrays of myofibrils (not shown).
138
SIMPSON ET AL. CARDIAC INTEGRINS
This is due in part to the binding redundancy that some integrin molecules exhibit. A variety of integrins can bind to the same matrix component. For example, the a3b1 and a5b1 as well as other integrins specifically interact and bind to fibronectin. Most cell types also simultaneously express a variety of different integrin complexes. Finally, integrins are not the sole physical connection between the cell surface and the external environment, sugar transferases appear to function in a similar fashion (39). Despite these complexities it is clear that integrin-mediated cell-matrix interactions play a critical role in mediating aspects of cell adhesion, migration, and differentiation. In particular, integrins of the b1 subfamily appear to mediate the cell-matrix interactions necessary for myofibrillogenesis (26), matrix remodeling (40), and cell sorting (41). The aligned myocyte cell culture system induces cardiac myocytes to express a predictable and nearly uniform cell shape. This makes it an ideal system in which to model how different constituents of the ECM or various integrins regulate the expression of a given phenotype. For example, function blocking antibodies directed against either the a1 or b1 integrin suppress the formation of the aligned phenotype (27). To further define the role of a1 subunit in the transduction of phenotypic information we have suppressed the expression of this integrin with antisense oligonucleotides (ODN) in aligned cultures of cardiac myocytes. In these experiments we used the antisense sequences composed of (59-CAGGCTCTTTTCTGTGGTGGA-39 [antisense ODN 1] and 59-GCTACTGTCGGGAACCTTATCT-39 [antisense ODN 2]) to suppress the expression of a1 integrin in cultured myocytes. As judged by 125I surface labeling and immunoprecipitation with specific antibodies these sequences reduce the total cell surface concentration of a1 integrin in cardiac myocytes up to 86% (Figure 3). In experiments with cultured neonatal cardiac myocytes we incubated freshly isolated cells with either antisense sequence 1 or 2, a nonsense ODN control, or a sense ODN control for 1 hour. The cells were then plated onto aligned collagen gels and incubated continuously in the presence of the appropriate ODN. Media was resupplemented at 12-hour intervals with fresh ODNs and the cells were refed at 24-hour intervals. After 72 hours of culture the cells incubated with nonsense or sense sequences of ODNs spread in parallel with the underlying matrix and expressed an elongated, rod-like phenotype (Figure 4). In contrast, incubation with antisense ODN 1 or ODN 2 disrupted cell spreading and the expression of the aligned phenotype (Figure 4). These data are consistent with our previous reports that the phenotypic information necessary to promote the formation of the aligned phenotype is carried primarily by the a1b1 integrin (27). We have also manipulated the profile of integrins present on the surface of aligned myocytes by infecting cultures with a replication deficient adenovirus designed to promote the overexpression of a2 or a5 integrin. Under ordinary circumstances neonatal cardiac myocytes express modest
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143
Figure 3. Antisense a1 integrin ODNs reduce the total cell surface concentration of endogenous a1 integrin. Myocytes were cultured for 48 hours in the absence or presence of various ODNs, rinsed in serum-free media, surface labeled with 125I, extracted and immunoprecipitated under non-reducing conditions with a1 integrin antibodies. Samples were normalized to total radioactive counts, separated by SDS gel electrophoresis on a 5–10% gel and prepared for autoradiography. Table depicts the relative optical density of the samples with respect to controls. Lane 1, control untreated cells, no ODN present. Lane 2, a1 integrin antisense ODN #1. Lane 3, a1 integrin antisense ODN #2. Lane 4, sense ODN for a1 integrin. Under non-reducing conditions precipitation with a1 integrin antibodies precipitates both the a and the non-covalently bound associated b subunit.
amounts of a5 on the plasma membrane and little or no detectable a2 (42). We plated replicate cultures of neonatal cardiac myocytes onto thin gels of aligned collagen for 96 hours. This interval of time allows the cells to fully assume the aligned phenotype and tissue-like pattern of organization. The cells were then pulse-labeled for 24 hours with 35S labeled methionine, rinsed in fresh media, and infected with the different adenovirus constructs (MOI of 20) for 4 hours. At intervals the cells were harvested and the concentration of contractile proteins or the amount of radioactivity remaining in different protein fractions was determined. We selected this paradigm because it allows us to measure the average metabolic profile of the cells over several days. Cardiac post-translational metabolism is very sensitive to changes in loading conditions (43,44), thus, interventions that stabilize or disrupt the physical interconnections between the cell and ECM can be expected to accelerate or suppress protein turnover.
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143
Figure 4. Suppression of a1 integrin expression with antisense ODNs inhibits the formation of the aligned phenotype. (A) cells treated with antisense ODNs directed against the a1 integrin. Many myocytes fail to spread and remain rounded even after 3–4 days of culture. Spread cells are elongated but not arrayed in parallel with one another or the underlying collagen fibrils (collagen is arrayed from right to left across the long axis of the micrograph). (B) cells treated with sense ODN and (C) cells treated with nonsense ODN. Non-specific effects of the charge modified ODNs are evident, compared with cells illustrated in Figure 2. Cultures are less dense, however, the cells express the elongated phenotype and are spread in parallel with one another and the underlying collagen fibrils.
In our experiments no significant differences were observed in total protein turnover in control uninfected cells, myocytes infected with a lac Z reporter gene (non-specific viral control), the a5 or the a2 integrin construct (Figure 5A). The overexpresion of the a5 construct was associated with a moderate acceleration in actin turnover (Figure 5B) and a substantial reduction in the total cellular concentration
SIMPSON ET AL. CARDIAC INTEGRINS
139
Figure 5. Overexpression of a2 and a5 in cultured myocytes. At various time points replicate cultures were extracted in equal volumes of SDS loading buffer and separated on 10% SDS gels. The gels were fixed and stained with Coomassie Blue. The relative amount of actin in the samples was determined by laser scanning densitometry, the relative amount of biosynthetically labeled actin was determined by autoradiography. Total protein turnover was determined by counting equal volumes of sample in a liquid scintillation counter [see (44) for a complete description of details and caveats]. No significant differences were noted in total protein turnover in control cells or cells infected with a2, a5, or Lac Z constructs (A). A modest acceleration in actin turnover was observed in cells overexpressing the a5 construct during the early phase of culture (B). The overexpression of a5 reduced the total cellular concentration of actin in the cultured cells (C). Data is expressed as the amount of radioactive counts (A) or optical density (B and C) remaining in the samples at different times as a percent time 0 (beginning of the chase interval, plus minus the standard error).
140
SIMPSON ET AL. CARDIAC INTEGRINS
(Figure 5C) of this contractile protein. This is the integrin subunit normally expressed at modest levels in the neonatal cardiac myocyte (42). We have previously reported that contractile protein turnover is intimately related to the state of myofibril structure (44,45). Interventions that disrupt myofibrillar structure are typically associated with an increase in contractile protein turnover. From these observations and the present results we predict that overexpressing a5 integrin in cardiac myocytes disrupts normal cell-ECM interconnections, reduces mechanical loading, and alters sarcomere architecture. Microscopic examination of cells infected with a5, but not a2, developed abnormalities in sarcomeric structure (not shown). Actin synthesis is also very sensitive to changes in mechanical load (45). It seems unlikely the modest acceleration in actin turnover can completely account for the depletion of this protein from the cultured cells in these experiments. Thus, it is probable that actin synthesis also is suppressed in cells overexpressing the a5 integrin. The overexpression of the a2 subunit may not have much of an effect in these experiments because this integrin is not normally expressed to any degree in cardiac myocytes (42). The myocytes may not recognize and/or process this subunit properly and, as a result its overexpression does not disrupt normal integrin-mediated signaling. In contrast, a5, which is normally expressed at low levels in these cells, may be capable of effectively competing with the endogenous cell-ECM-integrin signaling system when it is overexpressed. This interpretation is consistent with our observation that overcoating thin gels of aligned collagen with fibronectin, the substate for a5b1 integrin, disrupts the expression of the aligned phenotype by cardiac myocytes (27).
Integrin-Cytoskeletal Link Thus far we have considered how the ECM and its interconnections to the surface of cardiac myocytes regulates cardiac cell phenotype. Interconnections between the integrin and the cardiac cytoskeleton must also be intact for this linkage to be fully functional. Vinculin is a cytoskeletal protein found within the stroma of the Z-bands and as a component of the intercalated disks (46). It is also localized along the peripheral edges of the Z-bands immediately subjacent to the sarcolemma in structures called costameres. In the costamere, vinculin is a component of a multimolecular complex that anchors the Z-band to the integrin molecules that are concentrated along the lateral borders of the rodlike cardiac myocyte. Functional and structural assays have demonstrated that a physical continuum exists between the cytoskeleton, integrin molecules, and the extracellular matrix in these domains (47). This architectural arrangement allows for the propagation of contractile forces across the sarcolemma at the costamere. If this continuum is disrupted by selectively suppressing the expression of vinculin in cardiac myocytes plated onto thin gels of aligned collagen the cells fail to express the aligned phenotype (28). The inhibi-
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143
tion of normal vinculin expression in cardiac myocytes also suppresses cell spreading, the accumulation of contractile proteins, and the assembly of myofibrils.
Interactions Between Mechanical and Biochemical Signals Mechanical stimulation has been proposed to be a major determinant of cardiac cell structure and function (48). Physical perturbations, such as stretch, modulate cardiac gene expression (49,50), protein synthesis (51,52), and protein turnover (43–45). Mechanical activity also is necessary to promote the assembly of myofibrils, maintain the structural integrity of the contractile apparatus (44,45,53,54), and stabilize the distribution of integrins on the cell surface (55). We have already discussed how mechanical signals can be physically transmitted to the intracellular space by the cell-ECM-cytoskeletal linkage or biochemically communicated by phosphotyrosine cascades. Several recent studies have revealed how mechanical events can interact and work in concert with extracellular biochemical agents to regulate cardiac phenotype (56). For example, experiments with isolated cardiac myocytes and fibroblasts in tissue culture have shown that stretch induces the autocrine release of angiotensin II (AII). Applying this peptide to naive, unstretched cultures of myocytes accelerates protein synthesis and stimulates cardiac cell hypertrophy. In fibroblasts AII accelerates collagen expression (57). AII stimulates multiple intracellular secondary messenger systems in vitro (58). It is not known how (or if) AII interacts with other growth factors to direct cardiac development, however, since growth factors, such as PDGF and TGFb, stimulate the same secondary messenger pathways it is probable that there is cross-talk between these different systems. Several experiments indicate AII is not the sole carrier of stretchinduced hypertrophy in cardiac myocytes. AII does not stimulate protein synthesis to the same degree as stretch alone and it appears to have a differential effect on neonatal and adult cardiac myocytes (59). In addition, the intracellular signal pathways that are activated by angiotensin II are activated to some degree by stretch even in cardiac myocytes derived from angiotensin II knockout mice (60). However, the data indicate this peptide does represent a key component in signal pathways that are sensitive to stretch in the heart. We have examined the possible role of AII as a growth factor in cardiac development during early fetal life with the whole embryo culture system (61). This system makes it possible to examine the effects of various agents in the absence of confounding maternal influences. In these experiments embryos are removed from time pregnant rats at day 9.25 or 10.25 post coitus, dissected free of maternal tissues, and placed into rotating culture bottles. The embryos were then cultured in the presence or absence of AII or specific blockers of the AII receptors for 48 hours. In control ani-
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143
mals heart development proceeds over the time interval of 9.25 days 1 48 hours of culture from heart tube fusion to the late stages of cardiac looping (61). The interval of 10.25 days 1 48 hours of culture encompasses the middle stages of cardiac looping to the early stages of septation. Our experiments demonstrate that AII accelerates the accumulation of contractile proteins, promotes myofibrillar assembly, and increases overall cardiac growth (62). Stimulating cardiac growth with this peptide during the 9.25 6 48 hours of culture interval is also associated with a significant increase in cardiac inversions. Cardiac myocyte phenotype is also subject to modulation by a host of other soluble growth factors, cytokines, endocrine molecules, and neurohormonal agents (63–65). These agents chemically link the cardiac myocyte to the developmental and physiological state of the organism. Growth factors have been extensively studied as carriers of differentiation signals in the developmental process. These signal molecules elicit a diverse range of cellular responses. In developing cells these factors modulate mitogetic and migratory behavior (66) while in fully differentiated cells they can induce acute changes in cytoskeletal organization (67). Selected growth factors also stimulate the deposition of ECM components (68). The classical growth factors like PDGF, TGFb, FGF, and IGF have all been implicated in directing aspects of cardiac development and are also believed to assist in maintaining the normal adult phenotype (63–65). Developmental defects that have been ascribed to defective growth factor signaling range from gross dismorphogenesis of the heart to subtle functional deficits that may only be evident when the animal is physically challenged. Some growth factors, like PDGF, may exert their effects through multiple mechanisms. In development, this peptide functions as a classic growth factor and can modulate mitogenic activity (69) and function as a chemoattractant (70). When either PDGF ligand or its receptors are eliminated in mice during gene knockout experiments, the animals die in utero or during the very early hours of neonatal life (71,72). Knockout animals exhibit cardiovascular, renal, and hematological anomalies (72). In vitro experiments also suggest that PDGF can directly modulate the number of integrin receptors on the surface of cardiac myocytes (Terracio, unpublished observation) and fibroblasts (40). Altering the number of receptors on the cell surface could potentially affect a great number of developmental processes that we have already discussed, including the events that are sensitive to cell-matrix interactions and myofibrillar assembly (26,27). There is accumulating evidence to indicate that the cardiac interstitial compartment is a source as well as a repository, for peptide growth factors. Cardiac interstitial cells appear to produce peptides that function as growth factors to influence the growth, development, and maturation of the surrounding myocytes. Nonmuscle cells of the heart are also a source of secreted proteases that activate peptide growth factors and proenzymes bound to the extracellular matrix (73,74).
SIMPSON ET AL. CARDIAC INTEGRINS
141
Conclusions The integrin-ECM-cytoskeletal connection plays a critical role in regulating cardiac phenotype. Perturbations that alter any one component of this physical continuum can alter cell shape, protein metabolism, or myofibril order. In turn, interventions that alter cardiac phenotype must also have an impact on cardiac cell function. Mechanical forces play a critical role in directing many aspects of cardiac cell biology; however, accumulating evidence indicates that physical perturbations interact with soluble factors in the extracellular environment to control cardiac phenotype.
References 1. Olson E, Srivastava D. Molecular pathways controlling heart development. Science 1997;272:671–675. 2. Olson E. Things are developing in cardiology. Circ Res 1997;80:604– 606. 3. National Center for Health Statistics. Health United States DHHS Pub. No PHS90-1232. Washington, DC: US Govt Printing Office; 1990. 4. Healthy People 2000. DHHS Pub. No. PHS-94-50212. Washington, DC: US Govt Printing Office; 1991. 5. Lawrence PA, Struhl G. Morphogens, compartments and pattern. Lessons from drosophila. Cell 1996;85:951–956. 6. Raff M. Size control: the regulation of cell numbers in animal development. Cell 1996;86:173–175. 7. Bissell MJ, Hall HG, Parry G. How does the extracellular matrix direct gene expression? J Theor Biol 1982;99:31–68. 8. Streuli CH, Bissell MJ. Mammary epithelial cells, extracellular matrix and gene expression. Cancer Treatment Res 1991;53:365–381. 9. Aggeler J, Ward J, Blackie LM, Barcellos-Hoff MH, Streuli CH, Bissell MJ. Cytodifferentiation of mouse mammary epithelial cells cultured on a reconstituted basement membrane reveals striking similarities to development in vivo. J Cell Science 1993;99:407–417. 10. Howlett AR, Bissell MJ. The influence of tissue microenvironment (stroma and extracellular matrix) on the development and function of mammary epithelium. Epithelial Cell Bio 1993;2:79–89. 11. Sarber R, Hull B, Merrill C, Soranno T, Bell E. Regulation of proliferation of fibroblasts of low and high population doubling levels grown in collagen lattices. Mechanisms Ageing Dev 1981;17:107–117. 12. Kono T, Tanii T, Furukawa M, et al. Parallel arrangement, growth inhibition and cell cycle phase analysis of human dermal fibroblasts cultured in collagen lattice. J Dermatology 1990a;17:2–10. 13. Kono T, Tanii T, Furukawa M, et al. Cell cycle analysis of human dermal fibroblasts cultured on or in hydrated type I collagen lattices. Arch Dermatological Res 1990b;282:258–262. 14. Paye M, Nusgens BV, Lapiere CM. Modulation of cellular biosynthetic activity in the retracting collagen lattice. Eur J Cell Biol 1987; 45:44–50. 15. Robinson TF, Chohen-Gold L, Factor SM. Skeletal framework of mammalian heart muscle: arrangement of inter- and pericellular connective tissue structures. Lab Invest 1983;49:482–498. 16. Robinson TF, Factor SM, Capasso JM, Wittenberg BA, Blumenfeld OO, Seifter S. Morphology, composition, and function of struts between cardiac myocytes of rat and hamster. Cell Tiss Res 1987;249: 247–255. 17. Robinson TF, Geraci MA, Sonnenblick EH, Factor SM. Coiled perimysial fibers of papillary muscle in rat heart. Morphology, distribution, and changes in configuration. Circ Res 1988;63:577–592. 18. Robinson TF, Winegrad S. A variety of intercellular connections in heart muscle. J Mol Cell Cardiol 1981;13:185–195. 19. Borg TK. Development of the connective tissue network in the neonatal hamster heart. Anat Rec 1982;165:435–444.
142
SIMPSON ET AL. CARDIAC INTEGRINS
20. Borg TK, Terracio L. Interaction of the extracellular matrix with cardiac myocytes during development and disease. In: Robinson T, ed. Issues in Biomedicine. Basel: Karger Publishing, 1990;113–129.
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143
44.
21. Borg TK, Caulfield JB. Collagen in the heart. Tex Repts Biol Med 1979;39:321–333. 22. Caspari PG, Newcomb M, Gibson K, Harris P. Collagen in the normal and hypertrophied human ventricle. Cardiovasc Res 1977;11:554–558.
45.
23. Wickman-Coffelt J, Parmley WW, Mason DT. The cardiac hypertrophy process: analysis of factors determining pathological vs physiological development. Circ Res 1979;45:697–707.
46.
24. Weber KT, Janicki JS, Shroff SG, Chen RM, Bashey RI. Collagen remodeling of the pressure-overloaded, hypertrophied nonhuman primate myocardium. Circ Res 1988;62:757–765.
47.
25. Jalil JS, Doering CW, Janicki JS, Pick R, Shroff SG, Weber KT. Fibrillar collagen and myocardial stiffness in the intact hypertrophied rat left ventricle. Circ Res 1989;64:1041–1050.
48.
26. Hilenski LL, Xuehui M, Vinson N, Terracio L, Borg TK. The role of b1 integrin in spreading and myofibrillogenesis in neonatal rat cardiomyocytes in vitro. Cell Motil Cytoskel 1992;21:87–100.
49.
27. Simpson DG, Terracio L, Terracio M, Price RL, Turner DC, Borg TK. Modulation of cardiac myocyte phenotype in vitro by the composition and orientation of the extracellular matrix. J Cell Physiol 1994;161:89–105. 28. Shiraishi I, Simpson DG, Carver W, Hirozane T, Terracio L, Borg TK. Inhibition of vinculin with antisense oligodeoxynucleotides in fetal mouse cardiac myocytes. J Mol Cell Cardiol (in press).
50.
51.
29. Carver W, Terracio L. Integrin-Mediated Cell-Matrix Interactions in Heart Development and In Disease. Heart Failure 1993;8(6):255–263.
52.
30. Ingber D. Integrins as mechanochemical transducers. Curr Opin Cell Biol 1991;3:841–848.
53.
31. Juliano RL, Haskill S. Signal transduction from the extracellular matrix. J Cell Biol 1993;120:577–585. 32. Schwartz MA. Signaling by integrins: implications for tumorigenesis. Cancer Res 1993;53:1503–1506. 33. Watt FM, Kubler MD, Hotchin NA, Nicholson LJ, Adams JC. Regulation of keratinocyte terminal differentiation by integrin-extracellular matrix interactions. J Cell Sci 1993;106:175–182.
54.
55.
34. Mendrick DL, Kelly DM. Temporal expression of via-2 and modulation of its ligand specificity by rat glomerular epithelial cells in vitro. Lab Invest 1993;69:690–702.
56.
35. Carver W, Price RL, Raso DS, Terracio L, Borg TK. The distribution of b 1 integrin in the developing rat heart. J Histochem Cytochem 1994;42:167–175.
57.
36. Carver W, Bacro T, Borg TK, Terracio L. Expression and distribution of integrin a chains of the b1 subfamily in the developing rat heart. Dev Dynamics (in press). 37. Buck CA, Baldwin HS, Delisser H, et al. Cell adhesion receptors and early mammalian heart development: an overview. Comptes Rendus de l Academie des Sciences. Serie iii, Sciences de la Vie 1993; 316:838–859. 38. Song WK, Wang W, Foster RF, Biesler DAS, Kaufman SJ. H36-a7 is a novel integrin a chain that is developmentally regulated during skeletal myogenesis. J Cell Biol 1992;117:643–657. 39. Loeber CP, Runyan RB. A comparison of fibronectin, laminin, and galactosyltransferase adhesion mechanisms during embryonic cardiac mesenchymal cell migration in vitro. Devel Biol 1990;140:401–412. 40. Gullberg D, Tingstrom A, Thuresson A, et al. a1 integrin-mediated collagen gel contraction is stimulated by PDGF. Exp Cell Res 1990; 186:264–272. 41. Damsky CH, Moursi A, Zhou Y, Fisher SJ, Globus RK. The solid state of environment orchestrates embryonic development and tissue remodeling. Kidney Int 1997;51:1427–1433. 42. Terracio L, Rubin K, Balog E, et al. Expression of collagen binding integrins during cardiac development. Circ Res 1991;68:734–743. 43. Samarel AM, Spragia ML, Maloney V, Kamal SA, Engelmann GL.
58.
59.
60.
61. 62. 63.
64. 65.
Contractile arrest accelerates myosin heavy chain degradation in neonatal rat heart cells. Am J Physiol 1992;263(3):C642–652. Simpson DG, Sharp W, Terracio L, Price RL, Borg RK, Samarel AM. Mechanical regulation of cardiac protein turnover and myofibrillar structure. Am J Physiol 1996;270:C1075–C1087. Sharp WW, Terracio L, Borg TK, Samarel AM. Contractile activity modulates actin synthesis and turnover in cultured neonatal rat heart cells. Circ Res 1993;73(1):172–183. Terracio L, Simpson DG, Hilenski LH, et al. Distribution of vinculin in the Z-disk of striated muscle: analysis by laser scanning confocal microscopy. J Cell Physiol 1990;145(1):90–97. Danowski BA, Imanaka-Yoshida K, Sanger JM, Sanger JW. Costameres are sites of force transmission to the substratum in adult rat cardiomyocytes. J Cell Biol 1992;118:1411–1420. Simpson DG, Carver W, Borg TK, Terracio LT. The role of mechanical stimulation in the establishment and maintenance of muscle cell differentiation. Int Rev Cyto 1993;150:69–94. Komuro I, Katoh Y, Kaida T, et al. Mechanical loading stimulates cell hypertrophy and specific gene expression in cultured rat cardiac myocytes. Possible role of protein kinase C activation. J Biol Chem 1991; 266:1265–1268. Izumo S, Nadal-Ginard B, Mahdavi V. Protooncogene induction and reprogramming of cardiac gene expression produced by pressure overload. Proc Natl Acad Sci 1988;85:339–343. Morgan HE, Gordon EE, Chua BHL,Russo LA, Xenophontos XP. In: Dhalla NS, Sangel PK, Beamish RE, eds. Pathophysiology of Heart Disease. Amsterdam, NL: Martinus Nijhoff, 1987;93–98. Swynghedauw B. Biological adaptation of the myocardium to a permanent change in loading conditions. Basic Res Cardiol 1992;87:1–10. Kent RL, Uboh CE, Thompson EW, et al. Biochemical and structural correlates in unloaded and reloaded cat myocardium. J Mol Cell Cardiol 1985;17:153–165. Simpson DG, Decker ML, Clark WA, Decker RS. Contractile activity and cell-cell contact regulate myofibrillar organization in cultured cardiac myocytes. J Cell Biol 1993;123:323–326. Sharp WW, Simpson DG, Borg TK, Samarel AM, Terracio L. Contractile activity and external mechanical tension affect focal adhesion formation by cultured neonatal rat cardiac myocytes. Am J Physiol 1997;273:H546–H556. Sadoshima JI, Xu Y, Slater HS, Izumo I. Autocrine release of angiotensin II mediates stretch-induced hypertrophy of cardiac myocytes in vitro. Cell 1993;75:977–984. Ju H, Dixon IM. Effects of angiotensin II on collagen gene expression. Mol Cell Biochem 1996;163/164:231–237. Sadoshima J, Qui Z, Morgan JP, Izumo S. A II and other hypertrophic stimuli mediated by G protein-coupled receptors activate tyrosine kinase, mitogen-activated protein kinase, and 90 kD S6 kinase in cardiac myocytes. The critical role of Ca(12) dependent signaling. Circ Res 1995;76:1–15. Kent RL, McDermott PJ. Passive load and angiotensin II evoke differential responses in gene expression and protein synthesis in cardiac myocytes. Circ Res 1996;78:829–838. Nyui N, Tamura K, Mizumo K, et al. Stretch induced map kinase activation in cardiomyocytes of angiotensinogen defficient mice. Biochem Biophysic Res Commun 1997;235:36–41. Nakagawa M, Price RL, Chintanawongas C, et al. Analysis of heart development in cultured rat embryos. J Mol Cell Cardiol 1996;29:369–379. Price RL, Carver W, Simpson DG, et al. Effects of angiotensin II on cardiac growth and differentiation in early rat embryos. Dev Biol (in press). Schneider MD, McLellan WR, Black FM, Parker TG. Growth factors, growth factor response elements, and the cardiac phenotype. Basic Res Cardiol 1992;87:33–48. Long CS, Kariya K, Karns L, Simpson PC. Trophic factors for cardiac myocytes. J Hypertension 1990;8:S219–S224. Schneider MD, Parker TG. Cardiac growth factors. Progress Growth Factor Res 1991;3:1–26.
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143
66. Chen JD, Kim JP, Zhang K, et al. Epidermal growth factor (egf) promotes human keratinocyte locomotion on collagen by increasing the a 2 integrin subunit. Exp Cell Res 1993;209:216–223. 67. Chong LD, Traynor-Kaplan A, Bokoch GM, Schwartz MA. The small GTP-binding protein RHO regulates a phosphatidylinositol 4-phosphate 5-kinase in mammalian cells. Cell 1994;79:507–513. 68. Bienkowski RS, Gotkin MG. Growth factor mediated regulation of myocardial collagen gene expression. In: Eghbali-Webb M, ed. Molecular Biology of the Collagen Matrix of the Heart. Austin TX: RG Landes Co., 1995. 69. Bornfeldt KE, Raines RW, Graves LM, Skinner MP, Krebs EG, Ross R. PDGF. Distinct signal transduction pathways associated with migration vs. proliferation. Ann NY Acad Sci 1995;766:416–430.
SIMPSON ET AL. CARDIAC INTEGRINS
143
70. Abedi H, Zachary I. Signalling mechanisms in the regulation of vascular cell migration. Cardiovasc Res 1995;30:544–556. 71. Soriano P. Abnormal kidney development and hematolgocval disorders in PDGF B receptor mutant mice. Genes Dev 1994;8:1888–1896. 72. Leveen P, Pekney M, Gebre-Medhin S, Swolin B, Larsson E, Betsholtz C. Mice deficient for PDGF show renal, cardiovascular and hematological abnormalities. Genes Dev 1994;115:133–142. 73. Long CS, Hartgenesis WE, Simpson PC. A growth factor for cardiac myocytes is produced by cardiac non-myocytes. Cell Regulation 1991;2:1081–1095. 74. Borg KT, Burgess W, Terracio L, Borg TK. Expression of metalloproteases by cardiac myocytes and fibroblasts in vitro. Cardiovasc Pathol 1997;6:261–269.
144
SIMPSON ET AL. CARDIAC INTEGRINS
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143
Cardiac Integrins: The Ties That Bind David G. Simpson, Titus A. Reaves, Daw-tsun Shih, William Burgess, Thomas K. Borg, and Louis Terracio Department of Developmental Biology and Anatomy, School of Medicine, University of South Carolina, Columbia, South Carolina
11 An elaborate series of morphogenetic events must be precisely coordinated during development to promote the formation of the elaborate three-dimensional structure of the normal heart. In this study we focus on discussing how interconnections between the cardiac myocyte and its surrounding environment regulate cardiac form and function. In vitro experiments from our laboratories provide direct evidence that cardiac cell shape is regulated by a dynamic interaction between constituents of the extracellular matrix (ECM) and by specific members of the integrin family of matrix receptors. Our data indicates that phenotypic information is stored in the tertiary structure and chemical identity of the ECM. This information appears to be actively communicated and transduced by the a1b1 integrin molecule into an intracellular signal that regulates cardiac cell shape and myofibrillar organization. In this study we have assessed the phenotypic consequences of suppressing the expression and accumulation of the a1 integrin molecule in aligned cultures of cardiac myocytes. In related experiments we have examined how the overexpression of a2 and a5 integrin, integrins normally not present or present at very low copy number on the cell surface of neonatal cardiac myocytes, affect cardiac protein metabolism. We also consider how biochemical signals and the mechanical signals mediated by the integrins may converge on common intracellular signaling pathways in the heart. Experiments with the whole embryo culture system indicate that angiotensin II, a peptide that carries information concerning cardiac load, plays a role in controling cardiac looping and the proliferation of myofibrils during development. Cardiovasc Pathol 1998;7:135–143 © 1998 by Elsevier Science Inc.
The formation of the heart involves a series of well orchestrated morphogenetic events that must occur in a precisely coordinated sequential order. The developmental processes of commitment, migration, differentiation, and morphogenesis must all be integrated during fetal life to produce the elaborate three-dimensional structure of the heart (1,2). Deviations in this developmental program result in a variety of congenital malformations. These malformations range from relatively benign abnormalities in ventricular architecture to profound defects that result in the death of the developing fetus (3,4). The initial steps of early differentiation have come under intense investigation. However, a complete unManuscript received/accepted 20 November 1997. Presented as part of the International Society of Heart Research meeting, Vancouver, 1997 Address for correspondence: Louis Terracio, Department of Developmental Biology and Anatomy, University of South Carolina School of Medicine, Columbia, SC 29208; phone (803) 733-3369. Cardiovascular Pathology Vol. 7, No. 3, May/June 1998:135–143 1998 by Elsevier Science Inc. All rights reserved. 655 Avenue of the Americas, New York, NY 10010
derstanding of the regulatory processes that direct cardiac morphogenesis and the expression of cardiac cell phenotype has not been achieved. We focus this review on discussing how interconnections between the cell and its surrounding environment regulate cardiac form and function. Typically, the interconnections between the cell and its surrounding environment are regarded as a physical phenomenon, however, in many cases physical perturbations and the biochemical signals that link the heart to the surrounding interstitium and the other organ systems converge at a common intracellular second messenger pathway. We discuss some of potential sites where these signals converge. Fundamental principals of developmental biology, derived from elegant experimental paradigms in flies, nematodes, and fish are being applied to mammalian systems (5,6). Classical experiments demonstrated that developmental signals must be transmitted and received in a precise sequential fashion. As pointed out by Raff (6) the current
1054-8807/98/$19.00 PII S1054-8807(97)00123-3
136
SIMPSON ET AL. CARDIAC INTEGRINS
challenge is to understand how extracellular signals and intracellular events interact to regulate the differentiation of individual cells and the maturation of organ systems. The structural and functional profile of a cardiac myocyte undergoes a profound change during the developmental process. In the earliest stages of embryonic life the presumptive myocyte must migrate a considerable distance to assume its proper position within the condensing cell mass that will eventually form the heart. During this stage of development the presumptive cardiac myocyte has a nondescript cell shape, the capacity to proliferate, and lacks myofibrils. As development proceeds, the myocyte differentiates, assembles myofibrils, loses its capacity to divide, and assumes a rod-like cell shape. In order to fully understand how the distinctive phenotype of the adult cardiac myocyte is established and maintained, we must define the spatial and temporal sequence of events that direct cardiac development. The term phenotype is usually used to define the structural appearance, or shape, of a given cell type. However, in this discussion we use this term to describe both the structural and biochemical features that determine cell shape and function. We have adopted this convention because the structural characteristics of a cell are ultimately defined by its genetic potential and the biochemical processes that occur within the intracellular environment. From this definition it is clear that the evolution of a particular phenotype is the result of a dynamic interaction between the intrinsic genetic program of a cell and factors in the surrounding environment that regulate the extent to which this genetic program is implemented. Specific gene sequences can be regulated by the composition and arrangement of the surrounding extracellular matrix (ECM), mechanical signals, cell-to-cell communication, and a variety of soluble neurohumoral factors. It has become increasingly evident that a variety of cellular functions are intimately linked to the phenotype that a given cell type expresses (7,8). For example, in vitro many polarized secretory cells will not synthesize the appropriate complement of cellular proteins unless they are induced to “differentiate” and develop a basal, lateral, and apical surface (9,10). When fibroblasts are cultured in a three-dimensional matrix, instead of a planer surface, they express a more fusiform cell shape, display reduced proliferative capacity (11– 13), and alter their rate of collagen synthesis to reflect a more in vivo-like pattern (14). In the converse situation, if the normal structural phenotype of a cell is altered by an outside influence its ability to function may be drastically impaired. The scarring that occurs in association with an infarct compromises the structural integrity of the ventricular wall, modifies the conductive properties of the tissue, and alters the ability of the surrounding muscle cells to effectively contract.
Signals Associated with ECM In the normal adult myocardium the ECM is organized into a complex, three-dimensional network that envelops
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143
the individual cells of the heart (15–18). Selected components of this network are anchored to the surface of the myocytes through specific binding reactions (Figure 1). The ECM is produced largely by the fibroblasts, and its composition and arrangement is subject to change during development and normal aging. The biochemical and structural complexity of the ECM evolves gradually and in parallel with the developing heart (19–21). The chemical identity of the ECM also can be altered in disease states. Notably, during the onset of a pressure overload the ratio of type I to type III collagen is altered and the heart becomes less compliant (22–25). This has obvious consequences and alters cardiac contractile function on a global scale. Signals communicated from the ECM to the cell surface have been implicated in the regulation of biological events associated with cardiac development and the maintenance of the normal adult phenotype (26,27). In vitro experiments from our laboratories provide direct evidence that cardiac cell shape can be regulated by the composition and arrangement of the surrounding extracellular matrix. We have developed a substrate for cell culture composed of a thin film of aligned collagen fibrils (27). This matrix is prepared by applying a solution of ice cold, neutral collagen to a culture dish along a single, unique axis. The collagen solution is then drained from the culture substratum along the axis in which it was applied and the dish is then placed into a 378C incubator to polymerize the col-
Figure 1. Schematic of the ECM-integrin-cytoskeletal connection depicting the physical connection of the cytoskeleton to the ECM through the integrin complex. This continuum allows for the transmission of contractile forces from the myofibrils out into the ECM. Conversely, physical perturbations and other phenotypic signals can be communicated from the ECM to the intracellular space via the ECM-integrin-cytoskeletal connection. These external signals may be transmitted directly by mechanical ques or, indirectly, by biochemical signal cascades initiated by the integrin and/or molecules associated with the cytoplasmic tail of the integrin.
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143
lagen. The resulting matrix consists of collagen fibrils aligned along a common axis. When fetal (28) or neonatal cardiac myocytes (27) are cultured on this type of substrate they spread, as a population, in parallel with the fibrils of the underlying matrix. The individual cells of these cultures have an elongated, rod-like cell shape with myofibrils that are distributed in parallel with their long axis (Figure 2). In contrast, when neonatal cardiac myocytes are cultured on plastic, laminin, fibronectin, or a random film of collagen they express a flattened, stellate cell shape (27). Our experiments have demonstrated that a critical concentration of collagen is necessary to promote the expression of the rodlike phenotype. Applying too little, or too much, collagen to the dish disrupts the formation of the elongated cell shape and the tissue-like pattern of organization (27). Altering the profile of carbohydrates present on the collagen fibrils (unpublished observation, TA Reeves) or overcoating the collagen with fibronectin or laminin will also disrupt the expression of the aligned phenotype (27).
Integrin-ECM Connection The phenotypic information stored in the tertiary structure and chemical identity of the ECM is actively communicated to the cardiac myocyte. Constituents of the ECM are physically attached to the surface of these cells through specific interactions mediated by the integrin family of matrix receptors (29). The integrins are transmembrane molecules that are composed of an a and a b subunit. Various a and b chains combine to form receptors for different constituents of the ECM. Within the intracellular space, the cytoplasmic domains of the integrin complex interact with elements of the cytoskeleton (30), the internal scaffolding that supports
SIMPSON ET AL. CARDIAC INTEGRINS
137
and surrounds the myofibrils (Figure 1). Physical perturbations to the integrin complex are associated with the activation of phosphotyrosine cascades (30–32). Thus, external signals may be physically or biochemically communicated to the intracellular space by the integrin complex.
Integrin-ECM Signaling It is postulated that integrins transduce phenotypic information to cardiac myocytes in response to developmental events and disease processes that alter the composition or arrangement of the ECM. Integrins of the b1 subfamily are expressed in precise temporal and spatial patterns during embryogenesis in several organ systems including the integument (33) and kidney (34). This subfamily is also the predominate isoform of the myocardium (29). In embryonic life b1 mRNA and protein are expressed most abundantly in the developing trabeculae and cardiac cushions (35). As development proceeds, b1 protein accumulates in the ventricular myocytes and the leaflets of the valves. Preliminary evidence also indicates that the a chains of the integrin complex are expressed in very specific patterns in the developing heart (36). For instance, the integrin receptors for laminin and fibronectin are sequentially and regionally expressed in early heart development (37). With time this integrin subunit accumulates in the myocytes and subsequently persists in both the muscle and endothelial cells throughout embryonic and fetal life. The a7 chain of the integrin complex does not appear until late in fetal development and is restricted almost exclusively to the developing myocytes of the ventricular wall (38). It has been difficult to determine the functional role of individual integrin molecules during cardiac development.
Figure 2. Micrographs of the aligned myocyte phenotype. Cells exhibit an elongated cell shape and are distributed as a population along a common axis in a tissue-like pattern of organization (A). The underlying collagen fibrils (not visible) are arrayed across the long axis of the micrograph from right to left. The myofibrils of these rod-like cells are distributed in parallel with one another (B). Myocytes plated onto a non-aligned matrix of collagen express a stellate cell shape and random arrays of myofibrils (not shown).
138
SIMPSON ET AL. CARDIAC INTEGRINS
This is due in part to the binding redundancy that some integrin molecules exhibit. A variety of integrins can bind to the same matrix component. For example, the a3b1 and a5b1 as well as other integrins specifically interact and bind to fibronectin. Most cell types also simultaneously express a variety of different integrin complexes. Finally, integrins are not the sole physical connection between the cell surface and the external environment, sugar transferases appear to function in a similar fashion (39). Despite these complexities it is clear that integrin-mediated cell-matrix interactions play a critical role in mediating aspects of cell adhesion, migration, and differentiation. In particular, integrins of the b1 subfamily appear to mediate the cell-matrix interactions necessary for myofibrillogenesis (26), matrix remodeling (40), and cell sorting (41). The aligned myocyte cell culture system induces cardiac myocytes to express a predictable and nearly uniform cell shape. This makes it an ideal system in which to model how different constituents of the ECM or various integrins regulate the expression of a given phenotype. For example, function blocking antibodies directed against either the a1 or b1 integrin suppress the formation of the aligned phenotype (27). To further define the role of a1 subunit in the transduction of phenotypic information we have suppressed the expression of this integrin with antisense oligonucleotides (ODN) in aligned cultures of cardiac myocytes. In these experiments we used the antisense sequences composed of (59-CAGGCTCTTTTCTGTGGTGGA-39 [antisense ODN 1] and 59-GCTACTGTCGGGAACCTTATCT-39 [antisense ODN 2]) to suppress the expression of a1 integrin in cultured myocytes. As judged by 125I surface labeling and immunoprecipitation with specific antibodies these sequences reduce the total cell surface concentration of a1 integrin in cardiac myocytes up to 86% (Figure 3). In experiments with cultured neonatal cardiac myocytes we incubated freshly isolated cells with either antisense sequence 1 or 2, a nonsense ODN control, or a sense ODN control for 1 hour. The cells were then plated onto aligned collagen gels and incubated continuously in the presence of the appropriate ODN. Media was resupplemented at 12-hour intervals with fresh ODNs and the cells were refed at 24-hour intervals. After 72 hours of culture the cells incubated with nonsense or sense sequences of ODNs spread in parallel with the underlying matrix and expressed an elongated, rod-like phenotype (Figure 4). In contrast, incubation with antisense ODN 1 or ODN 2 disrupted cell spreading and the expression of the aligned phenotype (Figure 4). These data are consistent with our previous reports that the phenotypic information necessary to promote the formation of the aligned phenotype is carried primarily by the a1b1 integrin (27). We have also manipulated the profile of integrins present on the surface of aligned myocytes by infecting cultures with a replication deficient adenovirus designed to promote the overexpression of a2 or a5 integrin. Under ordinary circumstances neonatal cardiac myocytes express modest
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143
Figure 3. Antisense a1 integrin ODNs reduce the total cell surface concentration of endogenous a1 integrin. Myocytes were cultured for 48 hours in the absence or presence of various ODNs, rinsed in serum-free media, surface labeled with 125I, extracted and immunoprecipitated under non-reducing conditions with a1 integrin antibodies. Samples were normalized to total radioactive counts, separated by SDS gel electrophoresis on a 5–10% gel and prepared for autoradiography. Table depicts the relative optical density of the samples with respect to controls. Lane 1, control untreated cells, no ODN present. Lane 2, a1 integrin antisense ODN #1. Lane 3, a1 integrin antisense ODN #2. Lane 4, sense ODN for a1 integrin. Under non-reducing conditions precipitation with a1 integrin antibodies precipitates both the a and the non-covalently bound associated b subunit.
amounts of a5 on the plasma membrane and little or no detectable a2 (42). We plated replicate cultures of neonatal cardiac myocytes onto thin gels of aligned collagen for 96 hours. This interval of time allows the cells to fully assume the aligned phenotype and tissue-like pattern of organization. The cells were then pulse-labeled for 24 hours with 35S labeled methionine, rinsed in fresh media, and infected with the different adenovirus constructs (MOI of 20) for 4 hours. At intervals the cells were harvested and the concentration of contractile proteins or the amount of radioactivity remaining in different protein fractions was determined. We selected this paradigm because it allows us to measure the average metabolic profile of the cells over several days. Cardiac post-translational metabolism is very sensitive to changes in loading conditions (43,44), thus, interventions that stabilize or disrupt the physical interconnections between the cell and ECM can be expected to accelerate or suppress protein turnover.
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143
Figure 4. Suppression of a1 integrin expression with antisense ODNs inhibits the formation of the aligned phenotype. (A) cells treated with antisense ODNs directed against the a1 integrin. Many myocytes fail to spread and remain rounded even after 3–4 days of culture. Spread cells are elongated but not arrayed in parallel with one another or the underlying collagen fibrils (collagen is arrayed from right to left across the long axis of the micrograph). (B) cells treated with sense ODN and (C) cells treated with nonsense ODN. Non-specific effects of the charge modified ODNs are evident, compared with cells illustrated in Figure 2. Cultures are less dense, however, the cells express the elongated phenotype and are spread in parallel with one another and the underlying collagen fibrils.
In our experiments no significant differences were observed in total protein turnover in control uninfected cells, myocytes infected with a lac Z reporter gene (non-specific viral control), the a5 or the a2 integrin construct (Figure 5A). The overexpresion of the a5 construct was associated with a moderate acceleration in actin turnover (Figure 5B) and a substantial reduction in the total cellular concentration
SIMPSON ET AL. CARDIAC INTEGRINS
139
Figure 5. Overexpression of a2 and a5 in cultured myocytes. At various time points replicate cultures were extracted in equal volumes of SDS loading buffer and separated on 10% SDS gels. The gels were fixed and stained with Coomassie Blue. The relative amount of actin in the samples was determined by laser scanning densitometry, the relative amount of biosynthetically labeled actin was determined by autoradiography. Total protein turnover was determined by counting equal volumes of sample in a liquid scintillation counter [see (44) for a complete description of details and caveats]. No significant differences were noted in total protein turnover in control cells or cells infected with a2, a5, or Lac Z constructs (A). A modest acceleration in actin turnover was observed in cells overexpressing the a5 construct during the early phase of culture (B). The overexpression of a5 reduced the total cellular concentration of actin in the cultured cells (C). Data is expressed as the amount of radioactive counts (A) or optical density (B and C) remaining in the samples at different times as a percent time 0 (beginning of the chase interval, plus minus the standard error).
140
SIMPSON ET AL. CARDIAC INTEGRINS
(Figure 5C) of this contractile protein. This is the integrin subunit normally expressed at modest levels in the neonatal cardiac myocyte (42). We have previously reported that contractile protein turnover is intimately related to the state of myofibril structure (44,45). Interventions that disrupt myofibrillar structure are typically associated with an increase in contractile protein turnover. From these observations and the present results we predict that overexpressing a5 integrin in cardiac myocytes disrupts normal cell-ECM interconnections, reduces mechanical loading, and alters sarcomere architecture. Microscopic examination of cells infected with a5, but not a2, developed abnormalities in sarcomeric structure (not shown). Actin synthesis is also very sensitive to changes in mechanical load (45). It seems unlikely the modest acceleration in actin turnover can completely account for the depletion of this protein from the cultured cells in these experiments. Thus, it is probable that actin synthesis also is suppressed in cells overexpressing the a5 integrin. The overexpression of the a2 subunit may not have much of an effect in these experiments because this integrin is not normally expressed to any degree in cardiac myocytes (42). The myocytes may not recognize and/or process this subunit properly and, as a result its overexpression does not disrupt normal integrin-mediated signaling. In contrast, a5, which is normally expressed at low levels in these cells, may be capable of effectively competing with the endogenous cell-ECM-integrin signaling system when it is overexpressed. This interpretation is consistent with our observation that overcoating thin gels of aligned collagen with fibronectin, the substate for a5b1 integrin, disrupts the expression of the aligned phenotype by cardiac myocytes (27).
Integrin-Cytoskeletal Link Thus far we have considered how the ECM and its interconnections to the surface of cardiac myocytes regulates cardiac cell phenotype. Interconnections between the integrin and the cardiac cytoskeleton must also be intact for this linkage to be fully functional. Vinculin is a cytoskeletal protein found within the stroma of the Z-bands and as a component of the intercalated disks (46). It is also localized along the peripheral edges of the Z-bands immediately subjacent to the sarcolemma in structures called costameres. In the costamere, vinculin is a component of a multimolecular complex that anchors the Z-band to the integrin molecules that are concentrated along the lateral borders of the rodlike cardiac myocyte. Functional and structural assays have demonstrated that a physical continuum exists between the cytoskeleton, integrin molecules, and the extracellular matrix in these domains (47). This architectural arrangement allows for the propagation of contractile forces across the sarcolemma at the costamere. If this continuum is disrupted by selectively suppressing the expression of vinculin in cardiac myocytes plated onto thin gels of aligned collagen the cells fail to express the aligned phenotype (28). The inhibi-
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143
tion of normal vinculin expression in cardiac myocytes also suppresses cell spreading, the accumulation of contractile proteins, and the assembly of myofibrils.
Interactions Between Mechanical and Biochemical Signals Mechanical stimulation has been proposed to be a major determinant of cardiac cell structure and function (48). Physical perturbations, such as stretch, modulate cardiac gene expression (49,50), protein synthesis (51,52), and protein turnover (43–45). Mechanical activity also is necessary to promote the assembly of myofibrils, maintain the structural integrity of the contractile apparatus (44,45,53,54), and stabilize the distribution of integrins on the cell surface (55). We have already discussed how mechanical signals can be physically transmitted to the intracellular space by the cell-ECM-cytoskeletal linkage or biochemically communicated by phosphotyrosine cascades. Several recent studies have revealed how mechanical events can interact and work in concert with extracellular biochemical agents to regulate cardiac phenotype (56). For example, experiments with isolated cardiac myocytes and fibroblasts in tissue culture have shown that stretch induces the autocrine release of angiotensin II (AII). Applying this peptide to naive, unstretched cultures of myocytes accelerates protein synthesis and stimulates cardiac cell hypertrophy. In fibroblasts AII accelerates collagen expression (57). AII stimulates multiple intracellular secondary messenger systems in vitro (58). It is not known how (or if) AII interacts with other growth factors to direct cardiac development, however, since growth factors, such as PDGF and TGFb, stimulate the same secondary messenger pathways it is probable that there is cross-talk between these different systems. Several experiments indicate AII is not the sole carrier of stretchinduced hypertrophy in cardiac myocytes. AII does not stimulate protein synthesis to the same degree as stretch alone and it appears to have a differential effect on neonatal and adult cardiac myocytes (59). In addition, the intracellular signal pathways that are activated by angiotensin II are activated to some degree by stretch even in cardiac myocytes derived from angiotensin II knockout mice (60). However, the data indicate this peptide does represent a key component in signal pathways that are sensitive to stretch in the heart. We have examined the possible role of AII as a growth factor in cardiac development during early fetal life with the whole embryo culture system (61). This system makes it possible to examine the effects of various agents in the absence of confounding maternal influences. In these experiments embryos are removed from time pregnant rats at day 9.25 or 10.25 post coitus, dissected free of maternal tissues, and placed into rotating culture bottles. The embryos were then cultured in the presence or absence of AII or specific blockers of the AII receptors for 48 hours. In control ani-
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143
mals heart development proceeds over the time interval of 9.25 days 1 48 hours of culture from heart tube fusion to the late stages of cardiac looping (61). The interval of 10.25 days 1 48 hours of culture encompasses the middle stages of cardiac looping to the early stages of septation. Our experiments demonstrate that AII accelerates the accumulation of contractile proteins, promotes myofibrillar assembly, and increases overall cardiac growth (62). Stimulating cardiac growth with this peptide during the 9.25 6 48 hours of culture interval is also associated with a significant increase in cardiac inversions. Cardiac myocyte phenotype is also subject to modulation by a host of other soluble growth factors, cytokines, endocrine molecules, and neurohormonal agents (63–65). These agents chemically link the cardiac myocyte to the developmental and physiological state of the organism. Growth factors have been extensively studied as carriers of differentiation signals in the developmental process. These signal molecules elicit a diverse range of cellular responses. In developing cells these factors modulate mitogetic and migratory behavior (66) while in fully differentiated cells they can induce acute changes in cytoskeletal organization (67). Selected growth factors also stimulate the deposition of ECM components (68). The classical growth factors like PDGF, TGFb, FGF, and IGF have all been implicated in directing aspects of cardiac development and are also believed to assist in maintaining the normal adult phenotype (63–65). Developmental defects that have been ascribed to defective growth factor signaling range from gross dismorphogenesis of the heart to subtle functional deficits that may only be evident when the animal is physically challenged. Some growth factors, like PDGF, may exert their effects through multiple mechanisms. In development, this peptide functions as a classic growth factor and can modulate mitogenic activity (69) and function as a chemoattractant (70). When either PDGF ligand or its receptors are eliminated in mice during gene knockout experiments, the animals die in utero or during the very early hours of neonatal life (71,72). Knockout animals exhibit cardiovascular, renal, and hematological anomalies (72). In vitro experiments also suggest that PDGF can directly modulate the number of integrin receptors on the surface of cardiac myocytes (Terracio, unpublished observation) and fibroblasts (40). Altering the number of receptors on the cell surface could potentially affect a great number of developmental processes that we have already discussed, including the events that are sensitive to cell-matrix interactions and myofibrillar assembly (26,27). There is accumulating evidence to indicate that the cardiac interstitial compartment is a source as well as a repository, for peptide growth factors. Cardiac interstitial cells appear to produce peptides that function as growth factors to influence the growth, development, and maturation of the surrounding myocytes. Nonmuscle cells of the heart are also a source of secreted proteases that activate peptide growth factors and proenzymes bound to the extracellular matrix (73,74).
SIMPSON ET AL. CARDIAC INTEGRINS
141
Conclusions The integrin-ECM-cytoskeletal connection plays a critical role in regulating cardiac phenotype. Perturbations that alter any one component of this physical continuum can alter cell shape, protein metabolism, or myofibril order. In turn, interventions that alter cardiac phenotype must also have an impact on cardiac cell function. Mechanical forces play a critical role in directing many aspects of cardiac cell biology; however, accumulating evidence indicates that physical perturbations interact with soluble factors in the extracellular environment to control cardiac phenotype.
References 1. Olson E, Srivastava D. Molecular pathways controlling heart development. Science 1997;272:671–675. 2. Olson E. Things are developing in cardiology. Circ Res 1997;80:604– 606. 3. National Center for Health Statistics. Health United States DHHS Pub. No PHS90-1232. Washington, DC: US Govt Printing Office; 1990. 4. Healthy People 2000. DHHS Pub. No. PHS-94-50212. Washington, DC: US Govt Printing Office; 1991. 5. Lawrence PA, Struhl G. Morphogens, compartments and pattern. Lessons from drosophila. Cell 1996;85:951–956. 6. Raff M. Size control: the regulation of cell numbers in animal development. Cell 1996;86:173–175. 7. Bissell MJ, Hall HG, Parry G. How does the extracellular matrix direct gene expression? J Theor Biol 1982;99:31–68. 8. Streuli CH, Bissell MJ. Mammary epithelial cells, extracellular matrix and gene expression. Cancer Treatment Res 1991;53:365–381. 9. Aggeler J, Ward J, Blackie LM, Barcellos-Hoff MH, Streuli CH, Bissell MJ. Cytodifferentiation of mouse mammary epithelial cells cultured on a reconstituted basement membrane reveals striking similarities to development in vivo. J Cell Science 1993;99:407–417. 10. Howlett AR, Bissell MJ. The influence of tissue microenvironment (stroma and extracellular matrix) on the development and function of mammary epithelium. Epithelial Cell Bio 1993;2:79–89. 11. Sarber R, Hull B, Merrill C, Soranno T, Bell E. Regulation of proliferation of fibroblasts of low and high population doubling levels grown in collagen lattices. Mechanisms Ageing Dev 1981;17:107–117. 12. Kono T, Tanii T, Furukawa M, et al. Parallel arrangement, growth inhibition and cell cycle phase analysis of human dermal fibroblasts cultured in collagen lattice. J Dermatology 1990a;17:2–10. 13. Kono T, Tanii T, Furukawa M, et al. Cell cycle analysis of human dermal fibroblasts cultured on or in hydrated type I collagen lattices. Arch Dermatological Res 1990b;282:258–262. 14. Paye M, Nusgens BV, Lapiere CM. Modulation of cellular biosynthetic activity in the retracting collagen lattice. Eur J Cell Biol 1987; 45:44–50. 15. Robinson TF, Chohen-Gold L, Factor SM. Skeletal framework of mammalian heart muscle: arrangement of inter- and pericellular connective tissue structures. Lab Invest 1983;49:482–498. 16. Robinson TF, Factor SM, Capasso JM, Wittenberg BA, Blumenfeld OO, Seifter S. Morphology, composition, and function of struts between cardiac myocytes of rat and hamster. Cell Tiss Res 1987;249: 247–255. 17. Robinson TF, Geraci MA, Sonnenblick EH, Factor SM. Coiled perimysial fibers of papillary muscle in rat heart. Morphology, distribution, and changes in configuration. Circ Res 1988;63:577–592. 18. Robinson TF, Winegrad S. A variety of intercellular connections in heart muscle. J Mol Cell Cardiol 1981;13:185–195. 19. Borg TK. Development of the connective tissue network in the neonatal hamster heart. Anat Rec 1982;165:435–444.
142
SIMPSON ET AL. CARDIAC INTEGRINS
20. Borg TK, Terracio L. Interaction of the extracellular matrix with cardiac myocytes during development and disease. In: Robinson T, ed. Issues in Biomedicine. Basel: Karger Publishing, 1990;113–129.
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143
44.
21. Borg TK, Caulfield JB. Collagen in the heart. Tex Repts Biol Med 1979;39:321–333. 22. Caspari PG, Newcomb M, Gibson K, Harris P. Collagen in the normal and hypertrophied human ventricle. Cardiovasc Res 1977;11:554–558.
45.
23. Wickman-Coffelt J, Parmley WW, Mason DT. The cardiac hypertrophy process: analysis of factors determining pathological vs physiological development. Circ Res 1979;45:697–707.
46.
24. Weber KT, Janicki JS, Shroff SG, Chen RM, Bashey RI. Collagen remodeling of the pressure-overloaded, hypertrophied nonhuman primate myocardium. Circ Res 1988;62:757–765.
47.
25. Jalil JS, Doering CW, Janicki JS, Pick R, Shroff SG, Weber KT. Fibrillar collagen and myocardial stiffness in the intact hypertrophied rat left ventricle. Circ Res 1989;64:1041–1050.
48.
26. Hilenski LL, Xuehui M, Vinson N, Terracio L, Borg TK. The role of b1 integrin in spreading and myofibrillogenesis in neonatal rat cardiomyocytes in vitro. Cell Motil Cytoskel 1992;21:87–100.
49.
27. Simpson DG, Terracio L, Terracio M, Price RL, Turner DC, Borg TK. Modulation of cardiac myocyte phenotype in vitro by the composition and orientation of the extracellular matrix. J Cell Physiol 1994;161:89–105. 28. Shiraishi I, Simpson DG, Carver W, Hirozane T, Terracio L, Borg TK. Inhibition of vinculin with antisense oligodeoxynucleotides in fetal mouse cardiac myocytes. J Mol Cell Cardiol (in press).
50.
51.
29. Carver W, Terracio L. Integrin-Mediated Cell-Matrix Interactions in Heart Development and In Disease. Heart Failure 1993;8(6):255–263.
52.
30. Ingber D. Integrins as mechanochemical transducers. Curr Opin Cell Biol 1991;3:841–848.
53.
31. Juliano RL, Haskill S. Signal transduction from the extracellular matrix. J Cell Biol 1993;120:577–585. 32. Schwartz MA. Signaling by integrins: implications for tumorigenesis. Cancer Res 1993;53:1503–1506. 33. Watt FM, Kubler MD, Hotchin NA, Nicholson LJ, Adams JC. Regulation of keratinocyte terminal differentiation by integrin-extracellular matrix interactions. J Cell Sci 1993;106:175–182.
54.
55.
34. Mendrick DL, Kelly DM. Temporal expression of via-2 and modulation of its ligand specificity by rat glomerular epithelial cells in vitro. Lab Invest 1993;69:690–702.
56.
35. Carver W, Price RL, Raso DS, Terracio L, Borg TK. The distribution of b 1 integrin in the developing rat heart. J Histochem Cytochem 1994;42:167–175.
57.
36. Carver W, Bacro T, Borg TK, Terracio L. Expression and distribution of integrin a chains of the b1 subfamily in the developing rat heart. Dev Dynamics (in press). 37. Buck CA, Baldwin HS, Delisser H, et al. Cell adhesion receptors and early mammalian heart development: an overview. Comptes Rendus de l Academie des Sciences. Serie iii, Sciences de la Vie 1993; 316:838–859. 38. Song WK, Wang W, Foster RF, Biesler DAS, Kaufman SJ. H36-a7 is a novel integrin a chain that is developmentally regulated during skeletal myogenesis. J Cell Biol 1992;117:643–657. 39. Loeber CP, Runyan RB. A comparison of fibronectin, laminin, and galactosyltransferase adhesion mechanisms during embryonic cardiac mesenchymal cell migration in vitro. Devel Biol 1990;140:401–412. 40. Gullberg D, Tingstrom A, Thuresson A, et al. a1 integrin-mediated collagen gel contraction is stimulated by PDGF. Exp Cell Res 1990; 186:264–272. 41. Damsky CH, Moursi A, Zhou Y, Fisher SJ, Globus RK. The solid state of environment orchestrates embryonic development and tissue remodeling. Kidney Int 1997;51:1427–1433. 42. Terracio L, Rubin K, Balog E, et al. Expression of collagen binding integrins during cardiac development. Circ Res 1991;68:734–743. 43. Samarel AM, Spragia ML, Maloney V, Kamal SA, Engelmann GL.
58.
59.
60.
61. 62. 63.
64. 65.
Contractile arrest accelerates myosin heavy chain degradation in neonatal rat heart cells. Am J Physiol 1992;263(3):C642–652. Simpson DG, Sharp W, Terracio L, Price RL, Borg RK, Samarel AM. Mechanical regulation of cardiac protein turnover and myofibrillar structure. Am J Physiol 1996;270:C1075–C1087. Sharp WW, Terracio L, Borg TK, Samarel AM. Contractile activity modulates actin synthesis and turnover in cultured neonatal rat heart cells. Circ Res 1993;73(1):172–183. Terracio L, Simpson DG, Hilenski LH, et al. Distribution of vinculin in the Z-disk of striated muscle: analysis by laser scanning confocal microscopy. J Cell Physiol 1990;145(1):90–97. Danowski BA, Imanaka-Yoshida K, Sanger JM, Sanger JW. Costameres are sites of force transmission to the substratum in adult rat cardiomyocytes. J Cell Biol 1992;118:1411–1420. Simpson DG, Carver W, Borg TK, Terracio LT. The role of mechanical stimulation in the establishment and maintenance of muscle cell differentiation. Int Rev Cyto 1993;150:69–94. Komuro I, Katoh Y, Kaida T, et al. Mechanical loading stimulates cell hypertrophy and specific gene expression in cultured rat cardiac myocytes. Possible role of protein kinase C activation. J Biol Chem 1991; 266:1265–1268. Izumo S, Nadal-Ginard B, Mahdavi V. Protooncogene induction and reprogramming of cardiac gene expression produced by pressure overload. Proc Natl Acad Sci 1988;85:339–343. Morgan HE, Gordon EE, Chua BHL,Russo LA, Xenophontos XP. In: Dhalla NS, Sangel PK, Beamish RE, eds. Pathophysiology of Heart Disease. Amsterdam, NL: Martinus Nijhoff, 1987;93–98. Swynghedauw B. Biological adaptation of the myocardium to a permanent change in loading conditions. Basic Res Cardiol 1992;87:1–10. Kent RL, Uboh CE, Thompson EW, et al. Biochemical and structural correlates in unloaded and reloaded cat myocardium. J Mol Cell Cardiol 1985;17:153–165. Simpson DG, Decker ML, Clark WA, Decker RS. Contractile activity and cell-cell contact regulate myofibrillar organization in cultured cardiac myocytes. J Cell Biol 1993;123:323–326. Sharp WW, Simpson DG, Borg TK, Samarel AM, Terracio L. Contractile activity and external mechanical tension affect focal adhesion formation by cultured neonatal rat cardiac myocytes. Am J Physiol 1997;273:H546–H556. Sadoshima JI, Xu Y, Slater HS, Izumo I. Autocrine release of angiotensin II mediates stretch-induced hypertrophy of cardiac myocytes in vitro. Cell 1993;75:977–984. Ju H, Dixon IM. Effects of angiotensin II on collagen gene expression. Mol Cell Biochem 1996;163/164:231–237. Sadoshima J, Qui Z, Morgan JP, Izumo S. A II and other hypertrophic stimuli mediated by G protein-coupled receptors activate tyrosine kinase, mitogen-activated protein kinase, and 90 kD S6 kinase in cardiac myocytes. The critical role of Ca(12) dependent signaling. Circ Res 1995;76:1–15. Kent RL, McDermott PJ. Passive load and angiotensin II evoke differential responses in gene expression and protein synthesis in cardiac myocytes. Circ Res 1996;78:829–838. Nyui N, Tamura K, Mizumo K, et al. Stretch induced map kinase activation in cardiomyocytes of angiotensinogen defficient mice. Biochem Biophysic Res Commun 1997;235:36–41. Nakagawa M, Price RL, Chintanawongas C, et al. Analysis of heart development in cultured rat embryos. J Mol Cell Cardiol 1996;29:369–379. Price RL, Carver W, Simpson DG, et al. Effects of angiotensin II on cardiac growth and differentiation in early rat embryos. Dev Biol (in press). Schneider MD, McLellan WR, Black FM, Parker TG. Growth factors, growth factor response elements, and the cardiac phenotype. Basic Res Cardiol 1992;87:33–48. Long CS, Kariya K, Karns L, Simpson PC. Trophic factors for cardiac myocytes. J Hypertension 1990;8:S219–S224. Schneider MD, Parker TG. Cardiac growth factors. Progress Growth Factor Res 1991;3:1–26.
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143
66. Chen JD, Kim JP, Zhang K, et al. Epidermal growth factor (egf) promotes human keratinocyte locomotion on collagen by increasing the a 2 integrin subunit. Exp Cell Res 1993;209:216–223. 67. Chong LD, Traynor-Kaplan A, Bokoch GM, Schwartz MA. The small GTP-binding protein RHO regulates a phosphatidylinositol 4-phosphate 5-kinase in mammalian cells. Cell 1994;79:507–513. 68. Bienkowski RS, Gotkin MG. Growth factor mediated regulation of myocardial collagen gene expression. In: Eghbali-Webb M, ed. Molecular Biology of the Collagen Matrix of the Heart. Austin TX: RG Landes Co., 1995. 69. Bornfeldt KE, Raines RW, Graves LM, Skinner MP, Krebs EG, Ross R. PDGF. Distinct signal transduction pathways associated with migration vs. proliferation. Ann NY Acad Sci 1995;766:416–430.
SIMPSON ET AL. CARDIAC INTEGRINS
143
70. Abedi H, Zachary I. Signalling mechanisms in the regulation of vascular cell migration. Cardiovasc Res 1995;30:544–556. 71. Soriano P. Abnormal kidney development and hematolgocval disorders in PDGF B receptor mutant mice. Genes Dev 1994;8:1888–1896. 72. Leveen P, Pekney M, Gebre-Medhin S, Swolin B, Larsson E, Betsholtz C. Mice deficient for PDGF show renal, cardiovascular and hematological abnormalities. Genes Dev 1994;115:133–142. 73. Long CS, Hartgenesis WE, Simpson PC. A growth factor for cardiac myocytes is produced by cardiac non-myocytes. Cell Regulation 1991;2:1081–1095. 74. Borg KT, Burgess W, Terracio L, Borg TK. Expression of metalloproteases by cardiac myocytes and fibroblasts in vitro. Cardiovasc Pathol 1997;6:261–269.
144
SIMPSON ET AL. CARDIAC INTEGRINS
Cardiovasc Pathol Vol. 7, No. 3 May/June 1998:135–143