Induction of Coxsackievirus-Adenovirus–Receptor Expression During Myocardial Tissue Formation and Remodeling
Identification of a Cell-to-Cell Contact–Dependent Regulatory Mechanism
Background— The coxsackievirus-adenovirus receptor (CAR) was cloned as a receptor for both viruses, but its primary biological functions and regulatory mechanisms are unknown. CAR was low in healthy adult myocardium, whereas strong CAR reexpression was observed in human dilated cardiomyopathy. The molecular mechanisms of CAR induction in cardiomyocytes are unknown.
Methods and Results— We report on CAR regulation during development, CAR induction after myocardial infarction, and cell-to-cell contact–dependent CAR regulation in the rat. The high CAR expression during development in various organs decreased up to 190-fold after birth. After infarction resulting in severe cardiac dysfunction (dP/dtmax, −53%; dP/dtmin, −58%; left ventricular pressure, −45%), CAR was induced locally in cardiomyocytes of the infarct zone, where it was also expressed by capillary-like CD31+ structures and CD18+ interstitial cells, whereas it remained confined to subendothelial layers of arterioles and venules. In cultured cardiomyocytes, endothelin-1, cardiotrophin-1, leukemia-inhibiting factor, and cyclic stretch had no effect on CAR, whereas at high versus low cell density, CAR was suppressed up to 10-fold (P=0.006). Conditioned media from low- or high-density cardiomyocytes or cardiofibroblasts had no effect.
Conclusions— The locally confined CAR upregulation after infarction makes induction by various humoral factors unlikely, because cardiac dysfunction results in high activities of sympathetic and renin-angiotensin systems and cytokines. The cell culture experiments identify a cell-to-cell contact–dependent mechanism of CAR regulation. Further characterization of the signals linking cell-to-cell interactions to CAR gene expression may provide insight into mechanisms and functional consequences of the generalized CAR induction in dilated cardiomyopathy, and of its local induction after myocardial infarction.
Received August 2, 2002; revision received October 25, 2002; accepted November 6, 2002.
The coxsackievirus-adenovirus receptor (CAR) was cloned as a common receptor for both virus types, but its primary biological functions and regulatory mechanisms are unknown. CAR expression is highly variable among different tissues.1–4 Very low expression was found in normal adult human myocardium.5 Cell-to-cell interactions and the extracellular matrix play a key role in cell growth, differentiation and migration during embryonic development. Different cell adhesion molecules (CAMs) of the immunoglobulin superfamily are involved in these processes, and CAR is a new member of this family.1 In the brain, CAR seems to be involved in neural network formation.6 In Duchenne muscular dystrophy, the regeneration of muscle fibers is accompanied by CAR upregulation.3 CAR expression is modulated during myeloid and erythroid cell differentiation in the hematopoietic system.7 The function of CAR in the cardiovascular system is unknown, but strong myocardial induction of CAR was observed in human dilated cardiomyopathy (DCM). Human CAR was localized to sites of cell-to-cell contact similar to the intracellular targeting observed when recombinant CAR was expressed in cultured cardiomyocytes.5 CAR induction has also been described in rat autoimmune myocarditis.8 The molecular mechanisms of CAR induction in DCM are unknown. The biological significance of different CAR mRNA isoforms identified in various species, including humans, is also unknown.2 Whereas CAR induction occurs in vivo during normal development and in certain diseases, the discrimination of induction effects by humoral factors from those by cell-to-cell interactions requires extraction of the cell types under investigation to cell culture.
We report here first, on the differential regulation of two CAR mRNA isoforms,2 generated by alternative splicing,9 during rat development; second, on local and cell type–specific CAR induction after myocardial infarction (MI) in the rat; and third, on cell culture experiments that identify a cell-to-cell contact–dependent mechanism of CAR regulation in cardiomyocytes.
Animals, Embryos, and Tissues
Rats were from the Sprague-Dawley strain (Tierzucht Schoenwalde GmbH, Schoenwalde, Germany). Adult female rats used for vaginal plugging were 10 weeks old. The day of coitus was designated as day 0. Fetal tissues were collected from days 17 to 21 postcoitum. Tissues from newborn rats were collected from days 1 to 6 postpartum. Adult rats used for tissue collection were 3 months old. Tissues were frozen in liquid nitrogen immediately after decapitation.
Rat MI Model
Experiments were performed in male rats weighing 300 to 330 g with free access to water and standard chow under a 12-hour light/dark cycle. MI was induced by permanent ligation of the left anterior descending coronary artery.10 Hemodynamics were measured 6 hours, 6 days, and 3 weeks after MI with a Millar-tip catheter. Details are given in the online additions.
Cardiomyocyte Cultures, Cyclic Stretch, and Medium Transfer Experiments
Primary neonatal cardiomyocyte cultures were prepared from ventricular tissue of 1- to 3-day-old rats plated at different densities and subjected to various paracrine factors or to cyclic stretch as described.11–13 Details are given online.
Northern Blot Analyses and Quantitative Competitive RT-PCR and PCR Studies
Total RNA was isolated by use of the RNAclean and RNeasy test kits. The production of RNA and DNA standards, the procedures of quantitative competitive RT-PCR (qcRT-PCR) and PCR (qcPCR) reactions, the analysis of PCR products, and Northern analyses of cellular RNAs to determine GAPDH correction coefficients are described in the online additions.
Immunohistochemical Staining of CAR and Evaluation by Digital Image Analysis
Immunohistological detection and quantification of CAR in rat hearts after MI compared with sham-operated animals was performed as described previously.5 Details are given online.
Adenovector-Mediated Gene Transfer
Neonatal cardiomyocytes seeded on 6-well plates at 2×105 and 2.4×106 cells per well were transduced with 2×109 particles per well of a luciferase expressing adenovector (Ad5CMVluc) at 37°C for 20 minutes. Thereafter, cellular and adenovector (AdV) DNAs were isolated from the cells, and adenovector DNA uptake was quantified by qcPCR as described previously.14 Details are given online.
Details are given online.
CAR Regulation During Development and After MI
qRT-PCR quantification (online Figure A) of the expression levels of the two CAR mRNA isoforms in heart, brain, and skeletal muscle during development compared with adult rats showed up to 190-fold downregulation (Figure 1). In lung and liver (not shown), no downregulation was observed. Up to 12% of the total CAR mRNA was in the CAR2 isoform (online Figure B).
In contrast with the strong generalized cardiac CAR induction observed in human DCM,8 rat MI provides a model of local cardiac injury. We examined CAR expression in infarcted versus noninfarcted areas by using immunohistochemistry, digital image analysis, and confocal laser scanning microscopy (Figures 2 and 3⇓). CAR was analyzed 6 hours to 3 weeks after MI. After 6 days and 3 weeks, CAR induction was observed on cardiomyocytes of the infarct zone (Figure 2A), whereas cardiomyocyte CAR immunoreactivity in areas remote from the infarct was not elevated above the very low baseline of controls (online Figure C), which resembles the situation in normal human hearts.8 Myocardium remote from the infarct maintained low CAR expression, despite severe cardiac dysfunction in all post-MI animals: dP/dtmax had declined by 53% from 9540±430 mm Hg before MI to 4460±350 mm Hg, dP/dtmin by 58% from −7140±230 to −3360±200 mm Hg, and left ventricular pressure by 45% from 85.3±18.5 to 47.0±3.2 mm Hg, and left ventricular end-diastolic pressure had increased from 5.3±1.1 mm Hg before MI to 14.0±3.2 mm Hg (Table). In arterioles and venules, CAR expression was invariably confined to the subendothelial layers of the vascular wall in MI and control rats (Figure 2B and online Figure D), as in human hearts.5 However, CAR was detected on capillary-like CD31+ structures of the infarct zone (Figure 2C and online Figure E), which may represent angioneogenesis. Only small islands of CAR-positive interstitial cells were seen in normal myocardium, as in humans.5 In contrast, the infarct zone harbors a greatly increased number of CAR-positive interstitial cells (Figure 2D): mononuclear infiltrates and fibroblasts. Quantitative analyses of cardiac CAR are shown in Figure 3. The CAR area fraction (see Methods) assessed by digital image analysis over infarcted plus noninfarcted areas as a measure of total cardiac CAR rose 10-fold 3 weeks after MI (Figure 3). CAR induction was not immediate (no significant change after 6 hours), nor was it generalized all over the heart (no induction in areas remote from the infarct, whereas in the infarct zone, CAR area fraction rose ≈20-fold) (online Figure F).
Modulation of CAR Expression in Cardiomyocytes
In a search for factors influencing CAR in cardiomyocytes, cells were first incubated with endothelin-1 (ET-1), cardiotrophin-1 (CT-1), or leukemia inhibitory factor (LIF) or subjected to cyclic stretch. None of these factors had a significant effect on CAR mRNA or protein (online Figure G). In contrast, cell density exerted a marked influence, as documented both by CAR mRNA quantification by qcRT-PCR and by assessment of CAR function as a receptor. CAR1 mRNA abundance was 10-fold lower in high- versus low-density cells, whereas the CAR2 mRNA isoform decreased only 2-fold (Figure 4). CAR1 mRNA abundance in cardiomyocytes 48 hours after isolation ranged from 57% (lowest density) to only 6% (highest density) of the abundance in neonatal hearts. No change in abundance occurred because of different times in culture: abundance decreased rapidly after isolation of the cells to remain constant up to 72 hours of culture. Marker adenovirus uptake into cardiomyocytes was significantly higher at low versus high cell density, indicating functional regulation of CAR at the cell surface as well (online Figure G). A series of medium transfer experiments performed to identify or exclude involvement of a secreted factor in the cell density–dependent regulation of CAR is summarized in Figure 5. At high density, CAR1 mRNA was reduced (lanes 2 and 5). Admixture of cardiofibroblasts had no effect on CAR1 mRNA in low-density cardiomyocytes (lane 3). Neither transfer of culture medium conditioned by high-density cardiomyocytes to low-density cardiomyocyte cultures (lane 4) nor transfer of medium conditioned by low-density cardiomyocytes to high-density cardiomyocyte monolayers (lane 5) had any significant influence on CAR1 mRNA expression. Likewise, transfer of cardiofibroblast-conditioned medium had no effect (lane 6).
In summary, the high CAR expression levels observed during development decreased up to 190-fold in various organs after birth. CAR expression was induced locally in cardiomyocytes of the infarct zone but not in remote areas. CAR was also expressed by capillary-like CD31+ structures and interstitial cells of the infarct zone but was invariably confined to subendothelial wall layers of preformed vessels (arterioles and venules). The cell culture data indicate the existence of a cell-to-cell contact–dependent mechanism of CAR regulation in cardiomyocytes, whereas ET-1, CT-1, LIF, and cyclic stretch had no significant effect.
Induction of CAR in Cardiac Diseases
After human DCM displaying generalized myocardial CAR induction, experimental MI emerged as a cardiac disease associated with local CAR upregulation. Induction (defined as an increase in CAR immunoreactivity due to de novo CAR synthesis or exposure of immunoreactive sites) was observed in cardiomyocytes of the infarct zone but interestingly, not in remote myocardial areas, despite severe left ventricular dysfunction with its associated neurohumoral and cytokine activation. On preformed arterioles and venules, CAR expression remained confined to subendothelial wall layers even after infarction. Endothelial CAR induction has also not been detected in human terminal heart failure because of DCM and other causes.5 Subtle changes of CAR, which may nevertheless influence tight junction structure in vitro,15 cannot be excluded at the present level of light-microscopic resolution, but the lack of apparent endothelial induction distinguishes vascular regulation of CAR from that of numerous other CAMs, including members of the immunoglobulin superfamily.16 This seems to apply only for preformed vessels, however, whereas CAR expression was seen on capillary-like CD31+ structures of the infarct zone, which may represent angioneogenesis. Because completely different transcriptional regulation of genes occurs in developing versus mature vessels,17 the CAR gene may possibly respond to factors involved in early vascular development from endothelial tube formation to smooth muscle differentiation. Interestingly, the latter has been shown to be promoted by heterotypic cell-to-cell interactions.18,19 CAR may be involved in these early cell-to-cell contact events, whereas it appears not to be responsive to the various factors acting on mature vascular endothelia. Cardiac CAR induction is obviously not an unspecific response to cardiac injury, because it does not occur in heart failure of any origin and is locally confined to the infarct zone after MI, suggesting a physiological role in the complex processes of tissue remodeling in vivo. Although no direct in vitro model for these processes is available, future work with distinct isolated cell types in vitro may nevertheless identify functional alterations associated with CAR induction in those cells involved in vivo (cardiomyocytes and capillary-like and interstitial cells).
Local Versus Systemic Regulation of CAR
Together, the results from the developmental and cell biological studies and from the infarction model are consistent with the hypothesis that CAR expression is regulated locally rather than systemically. The CAR regulatory mechanism identified here (downregulation by high cardiomyocyte density) apparently requires direct cell-to-cell communication. The lack of any effect on CAR expression of factors acting, for example, through heptahelical G protein–coupled (endothelin-1), gp130/LIFRβ receptor (CT-1, LIF), or Ptd-Ins-3 kinase/Akt pathway (cyclic stretch)20 distinguishes regulation of CAR from that of other fetal proteins in the heart (eg, atrial natriuretic factor). Because few of the factors/receptors known to affect cardiac structure and function have been examined here, the existence of other signaling systems regulating CAR expression cannot be excluded. However, in addition to the cardiomyocyte culture studies, the CAR patterns observed after MI also make unlikely any significant regulation of CAR by numerous other factors. In the post-MI animals with severe cardiac dysfunction and its associated strong neurohumoral activation and high circulating levels of catecholamines, ET-1, angiotensin II, and cytokines (tumor necrosis factor-α, interleukin-6), cardiac CAR induction still remained confined to the infarct zone. Independent of this observation, severe heart failure per se was not associated with myocardial CAR induction in humans, either.5 If the aforementioned factors contributed significantly to induction, it should occur in all parts of the failing hearts and in severe heart failure of any origin. High local concentrations of factors not studied here may contribute to CAR regulation, but to the best of our knowledge, no individual factor of that kind has yet been identified. Myocardial CAR induction has been described in rat autoimmune myocarditis and by concanavalin A–stimulated lymphocyte supernatants, whereas no individual CAR-inducing cytokine was identified.8 Inflammatory cytokines or other immune factors cannot play any role in the cardiomyocyte density–dependent CAR regulation in vitro described here, and no pathological inflammation occurs during normal development. However, during autoimmune myocarditis and after MI, complex remodeling processes associated with inflammatory cell infiltrations may also influence CAR independently of the mechanisms discussed below.
Cell-to-Cell Contact–Dependent Mechanism of CAR Regulation
The observation of strong CAR downregulation at the mRNA level at high cardiomyocyte densities suggests cell-to-cell contact–dependent signaling as one possible regulatory mechanism of CAR gene transcription. Indeed, the cellular targeting of recombinant CAR in cardiomyocytes and the CAR expression patterns in human DCM5 indicate that CAR does fulfill one important requirement for molecules involved in cell-to-cell contact–dependent signaling: ie, aggregation at sites of cell-to-cell contacts.21 In addition, the medium transfer experiments summarized in Figure 5 indicate that the regulation does not occur via diffusible intermediates. Cell-to-cell contact–dependent signals (in contrast to paracrine signals requiring diffusible intermediates such as norepinephrine, ET-1, angiotensin-II, CT-1, LIF, tumor necrosis factor-α, interleukin-6) have been shown to play a key role in fine patterning during embryogenesis, whereas paracrine signaling allows patterning of large embryonic fields.21 Among the molecules involved in cell-to-cell contact–dependent signaling, certain CAMs appear to trigger intracellular signals after homophilic CAM-to-CAM binding, instead of interactions between membrane-bound but distinct ligand-receptor pairs (juxtacrine signaling).21 In vitro, the human CAR D1 domain does form dimers.22 If membrane-bound CAR on distinct cells does also form homodimers in vivo, triggering intracellular signals, then the two isoforms of CAR may be involved in distinct biological processes, depending on different signaling functions of their intracellular domains and on cell type.23 Cell-to-cell contact–mediated, CAR-dependent signals modulating cell cycle regulators in cancer cells have been proposed to explain the growth-inhibitory effect of CAR on human bladder cancer.24 There is striking similarity between the tyrosine phosphatase DEP-125 and CAR: DEP-1 is regulated by cell density in multiple cell types. Transfection of DEP-1 is growth inhibitory. Furthermore, other viral receptors have recently been shown to be membrane-bound phosphatases.
In conclusion, the present and previous studies5,8 in cultured cells and in normal and diseased organs in vivo are first steps toward clarification of CAR regulatory mechanisms and functions in the various cell types interacting with each other and with humoral factors in the cardiovascular system. Starting from the observation of CAR expression on capillary-like structures after MI in vivo, additional in vitro investigations are necessary to characterize CAR regulation during angioneogenesis and its possible role in the process. Similarly, our cardiomyocyte experiments in vitro need not indicate that CAR is regulated by geometrical distance between cardiomyocytes (“density”) in vivo, too. However, they show the existence of cell-to-cell contact–dependent and possibly CAR-to-CAR interaction–based signals, which may be disturbed in vivo by factors requiring further studies. Anti-CAR antibodies and soluble CAR isoforms should help to characterize these signaling pathways. In the case of DCM, the strong generalized CAR induction observed may be an unimportant consequence of the basic disease process (notwithstanding its relevance for virus uptake), but it may also be causally involved in this process by either aggravating or alleviating it, acting as a trigger for functional changes in the myocardium.
This work was supported by a Heisenberg fellowship (WP378/2-1,2-2) and by a research grant (WP378/6-1) from the Deutsche Forschungsgemeinschaft to Dr Poller. Kerstin Hinze was supported by a grant from the Freie Universität, Berlin, Germany. We thank Ulrike Siemetzki, Norbert Hinz, and Frank Spillmann for excellent technical assistance.
An expanded Methods section with additional figures is available in the online-only Data Supplement at http://www.circulationaha.org.
Tomko R, Xu R, Philipson L. HCAR and MCAR: the human and mouse cellular receptors for subgroup C adenoviruses and group B coxsackieviruses. Proc Natl Acad Sci U S A. 1997; 94: 3352–3356.
Bergelson J, Krithivas A, Celi L, et al. The murine CAR homolog is a receptor for coxsackie B viruses and adenoviruses. J Virol. 1998; 72: 415–419.
Noutsias M, Fechner H, de Jonge H, et al. Human coxsackie-adenovirus receptor is colocalized with integrins alpha(v)beta(3) and alpha(v)beta(5) on the cardiomyocyte sarcolemma and upregulated in dilated cardiomyopathy: implications for cardiotropic viral infections. Circulation. 2001; 104: 275–280.
Ito M, Kodama M, Masuko M, et al. Expression of coxsackievirus and adenovirus receptor in hearts of rats with experimental autoimmune myocarditis. Circ Res. 2000; 86: 275–280.
van Kesteren C, Saris JJ, Dekkers DH, et al. Cultured neonatal rat cardiac myocytes and fibroblasts do not synthesize renin or angiotensinogen: evidence for stretch-induced cardiomyocyte hypertrophy independent of angiotensin II. Cardiovasc Res. 1999; 43: 148–156.
Cohen C, Shieh J, Pickles R, et al. The coxsackievirus and adenovirus receptor is a transmembrane component of the tight junction. Proc Natl Acad Sci U S A. 2001; 98: 15191–15196.
Noutsias M, Seeberg B, Schultheiss H, et al. Expression of cell adhesion molecules in dilated cardiomyopathy: evidence for endothelial activation in inflammatory cardiomyopathy. Circulation. 1999; 99: 2124–2131.
Oettgen P. Transcriptional regulation of vascular development. Circ Res. 2001; 89: 380–388.
Hirschki K, Rohovsky S, D’Amore P. PDGF, TGF-β, and heterotypic cell-cell interactions mediate endothelial cell-induced recruitment of 1oT1/2 cells and their differentiation to a smooth muscle fate. J Cell Biol. 1998; 141: 805–814.
Hirschki K, Rohovsky S, Beck L, et al. Endothelial cells modulate the proliferation of mural cell precursors via platelet-derived growth factor-BB and heterotypic cell contact. Circ Res. 1999; 84: 298–305.
Carson S, Hobbs J, Tracy S, et al. Expression of the coxsackievirus and adenovirus receptor in cultured human umbilical vein endothelial cells: regulation in response to cell density. J Virol. 1999; 73: 7077–7079.
Okegawa T, Pong R-C, Li Y, et al. The mechanism of the growth-inhibitory effect of coxsackie and adenovirus receptor (CAR) on human bladder cancer: a functional analysis of CAR protein structure. Cancer Res. 2001; 61: 6592–6600.
Kovalenko M, Denner K, Sandström J, et al. Site-selective dephosphorylation of the platelet-derived growth factor beta-receptor by the receptor-like protein-tyrosine phosphatase DEP-1. J Biol Chem. 2000; 275: 16219–16226.