Human Coxsackie-Adenovirus Receptor Is Colocalized With Integrins αvβ3 and αvβ5 on the Cardiomyocyte Sarcolemma and Upregulated in Dilated Cardiomyopathy
Implications for Cardiotropic Viral Infections
Background The coxsackievirus and adenovirus receptor (CAR) was identified as a common cellular receptor for both viruses, but its biological and pathogenic relevance is uncertain. Knowledge of CAR localization in the human cardiovascular system is limited but important with respect to CAR-dependent viral infections and gene transfer using CAR-dependent viral vectors.
Methods and Results Explanted failing hearts from 13 patients (8 with dilated cardiomyopathy [DCM] and 5 with other heart diseases [non-DCM]) and normal donor hearts (n=7) were investigated for the expression levels and subcellular localization of CAR and the adenovirus coreceptors αvβ3 and αvβ5 integrins. CAR immunoreactivity was very low in normal and non-DCM hearts, whereas strong CAR signals occurred at the intercalated discs and sarcolemma in 5 of the 8 DCM hearts (62.5%); these strong signals colocalized with both integrins. In all hearts, CAR was detectable in subendothelial layers of the vessel wall, but not on the luminal endothelial surface, and on interstitial cells. Human CAR (hCAR) expressed in rat cardiomyocytes was targeted to cell-cell contacts, which resembled CAR localization in DCM hearts and resulted in 15-fold increased adenovirus uptake.
Conclusions Low hCAR abundance may render normal human myocardium resistant to CAR-dependent viruses, whereas re-expression of hCAR, such as that observed in DCM, may be a key determinant of cardiac susceptibility to viral infections. Asymmetric expression of hCAR in the vessel wall may be an important determinant of adenovirus tropism in humans. hCAR subcellular localization in human myocardium and hCAR targeting to cell-cell contacts in cardiomyocyte cultures suggest that hCAR may play a role in cell-cell contact formation.
Received November 29, 2000; revision received April 4, 2001; accepted April 17, 2001.
The coxsackie-adenovirus receptor (CAR), a 46-kDa transmembrane protein, was cloned as a common cellular receptor for both viruses.1,2⇓ Despite detailed knowledge about the molecular structure and virus interaction of CAR,3,4⇓ its biological and possible pathogenic relevance is uncertain. CAR expression, which is relevant for both viral pathogenesis and gene therapy with CAR-dependent vectors, is highly variable among different tissues in various species.2,5,6⇓⇓ Knowledge of CAR localization in the human cardiovascular system at the cellular and subcellular level is limited; however, it is important for understanding the molecular mechanisms of human coxsackievirus7 and adenovirus infections8 and the tropism of CAR-dependent viral vectors. Recent work on experimental autoimmune myocarditis in the rat model showed CAR upregulation during myocardial inflammation.9
We report here on cardiac CAR expression patterns in patients with heart failure due to dilated cardiomyopathy (DCM) or other heart diseases; our data suggest that CAR is dynamically regulated in the human cardiovascular system. We found a low abundance of CAR in normal human myocardium but high cardiac expression of CAR in DCM hearts. Implications for cardiotropic viral infections and therapeutic cardiac gene delivery using CAR-dependent vectors, which has been investigated as a novel approach to the treatment of cardiac disorders,10–14⇓⇓⇓⇓ are discussed. We also found that recombinant human CAR (hCAR) expressed in neonatal rat cardiomyocytes was targeted to cell-cell contact sites in a way reminiscent of the CAR expression patterns in DCM hearts, which showed hCAR localized to intercalated discs and sarcolemma. These data, taken together with previous reports that CAR is highly expressed during embryogenesis and is involved in cell-cell recognition,9,15⇓ suggest that CAR may be a “pathfinder” protein which is transiently expressed during embryogenesis to promote cell-cell contact formation, that it is strongly downregulated in normal adult human myocardium, and that it may become re-expressed in certain diseases.
The clinical data concerning the investigated patients are summarized in the Table. The explanted hearts of 13 patients with terminal heart failure who were undergoing heart transplantation (HTX) were studied. In 8 patients, heart failure was due to DCM; in 3 patients, it was due to coronary artery disease with end-stage ischemic cardiomyopathy; in 1 patient, it was due to familial hypertrophic nonobstructive cardiomyopathy; and in 1 patient, it was due to transposition of the great vessels and atrial and ventricular septal defects. Age at HTX in the 8 DCM patients (1 woman and 7 men) was 52.3±10.2 years (range, 31.6 to 65.4 years), and left ventricular ejection fraction was 18.9±3.9% (range, 12% to 25%). Age at HTX in the 5 non-DCM patients (1 woman and 4 men) was 40.9±14.4 years (range, 26.7 to 60.4 years), and ejection fraction was 19.6±3.2% (range, 15% to 23%). All patients were in New York Heart Association (NYHA) class 3 to 4 at the time of HTX. Biopsies from 7 explanted donor hearts scheduled for HTX were used as controls (donor group).
Immunohistochemistry, Immunofluorescence, Digital Image Analysis, and Microscopy
Tissue samples were removed from the ventricular septum of cardiomyopathic hearts after explantation. From the donor hearts, endomyocardial biopsies were removed using a bioptome before implantation. All patients gave informed consent. Detailed descriptions of the immunohistochemical, immunofluorescence, digital image analysis, and confocal laser scanning microscopy procedures performed on human tissues and cultured cardiomyocytes can be found in the Data Supplement (located at www.circulationaha.org).
Cardiomyocyte Cultures and Gene Transfer
Cloning a recombinant adenoviral hCAR vector and neonatal rat cardiomyocyte cultures were performed essentially as described previously.10,16–18⇓⇓⇓ A detailed description can be found in the Data Supplement.
Details of the statistical procedures performed are in the Data Supplement.
Expression Patterns of Myocardial CAR Expression in Human Hearts
The clinical characteristics of the patients investigated are summarized in the Table. There were no significant differences between DCM and non-DCM patients with respect to sex, age at HTX, ejection fraction, or NYHA class. When normal human hearts or non-DCM hearts were immunostained for CAR, there was no detectable immunoreactivity on cardiomyocyte sarcolemma or intercalated discs. There was clear immunostaining of interstitial cells (Figure 1C), which accounts for the non-zero CAR “area fraction” (defined in the Data Supplement), as determined both in the non-DCM and the donor group (Figure 2). CAR was also detected on vascular smooth muscle cells, but there was no detectable CAR immunoreactivity on the luminal surface of cardiac microvascular endothelium (Figure 1D), with no significant differences between the donor, non-DCM, and DCM group. In sharp contrast, there was clear CAR immunostaining on both cardiomyocyte sarcolemma and intercalated discs in 5 of the 8 DCM hearts (62.5%). Conventional immunohistochemistry of this finding is shown in Figure 1A.
The staining of the cardiomyocyte sarcolemma was evident whether the tissues were cross, longitudinally, or traversely cut. When omitting the anti-CAR antibody from the staining procedure, no immunoreactivity was discernible (Figure 1B). Quantitation of CAR area fraction (Figure 2) showed significantly (P<0.01) higher values in the DCM group (0.0145±0.0056) than in the non-DCM (0.0013±0.0006) or donor groups (0.0004±0.0006). Confocal laser scanning microscopy with single and triple staining for CAR, the integrins αvβ3 and αvβ5, and CD31 revealed that the CAR immunoreactivity prominent on the intercalated discs and the cardiomyocyte sarcolemma (Figure 3A) was colocalized with the integrins αvβ3 and αvβ5 (Figures 3A and B), which are involved in adenovirus internalization, but not with CD31, which was expressed only by endothelial cells. Furthermore, colocalization of CAR, αvβ3, and αvβ5 was also observed on interstitial cells, which again were devoid of CD31 reactivity.
In summary, there was significant abundance of myocardial CAR expression in 62.5% of the DCM patients, in whom CAR was colocalized with the integrins αvβ3/5 at the intercalated discs, the sarcolemma, and interstitial cells. CAR was absent from the luminal surface of cardiac microvessels but present in vascular smooth muscle cells and interstitial cells, without significant differences among the 3 groups.
Cellular Targeting of Human CAR in Neonatal Cardiomyocytes
Because normal and non-DCM myocardium showed very low CAR immunoreactivity, re-expression in the DCM myocardium could possibly follow some embryonic expression pattern. Following this hypothesis, we investigated the cellular targeting of hCAR within neonatal cardiomyocytes using the model of recombinant hCAR expression in neonatal rat cardiomyocytes. Figures 4C and 4D show the distinct hCAR immunoreactivity patterns observed in the cardiomyocyte cultures. No staining of untransfected rat cardiomyocytes was detected (Figures 4A and 4B), but hCAR could be detected on transfected cells (Figures 4C through 4F). There was strong targeting of hCAR to cell-cell contact sites between cardiomyocytes. This type of targeting was also seen when occasional noncardiomyocytes were present in the cardiomyocyte cultures. Figures 4E and 4F show finger-like structures emanating from a noncardiomyocyte (cardiac fibroblast-like cell) to 2 different cardiomyocytes. Again, CAR immunoreactivity was sharply focused to cell-cell contact sites, which are very small in this case. To further characterize the nature of the described cell-cell contact sites, we looked for possible colocalization and coexpression of CAR and the focal adhesion-associated protein paxillin and the gap junction-associated protein connexin43, respectively. However, those proteins were neither spatially nor temporally coexpressed; they follow clearly distinct expression patterns (data not shown).
Enhancement of Adenoviral Infection by CAR Overexpression
By overexpressing human CAR in the cardiomyocytes, adenovirus attachment to and uptake into the cells was strongly enhanced (Figure I in Data Supplement), suggesting that CAR is a key determinant of cardiac susceptibility to adenovirus infections, with the coreceptors αvβ3 and αvβ5 not being a limiting factor.
Highly Variable CAR Expression Patterns in Human Hearts
CAR is used by 2 unrelated viruses as a common cellular receptor,1,2⇓ but its physiological role at the cellular level and its possible pathogenic relevance are currently unknown. Our findings of highly variable CAR expression patterns in normal versus diseased human myocardium and of specific cellular CAR targeting in cardiomyocyte cultures lead to the following 2 hypotheses: (1) the low abundance of CAR in normal and non-DCM human hearts may render them rather resistant to infection by CAR-dependent viruses, whereas CAR upregulation may be an important predisposing factor for CAR-dependent viral infections of the human myocardium; and (2) CAR may play important physiological and biological roles in cell-cell contact formation between cardiomyocytes, consistent with the findings both in human tissues and in cardiomyocyte cultures.
Possible Causes and Pathogenic Consequences of Altered CAR Expression
Although interstitial cells expressed CAR in both DCM and control hearts, CAR expression in cardiomyocytes was absent from donor hearts and hearts from patients with terminal heart failure entities other than DCM; instead, it was significantly induced in the cardiomyocytes of DCM hearts. The presence of CAR in cardiomyocytes was observed in 62.5% of the investigated DCM hearts. This staining pattern was demonstrated by both immunohistochemical and immunofluorescence techniques. Technical pitfalls, such as cross-reactivity of the anti-mouse antibody applied in immunohistochemistry or cross-binding fluorophores in immunofluorescence, were excluded: tissues stained omitting the anti-CAR antibody did not display any immunoreactivity, no cross-staining of the fluorescent dyes was observed when modifying the sequence of staining, and the anti-CD31 antibody clearly depicted exclusive staining of endothelium and no interference with CAR. Whether there is “acquired” up-regulation of CAR from a low basal level or if some individuals maintain permanent high CAR levels on a genetic basis cannot be decided thus far, but the latter could predispose such individuals for cardiac disease.
Given the limited number of hearts examined, no definitive generalization is currently possible about the possible disease specificity of this phenomenon. Interestingly, however, although myocardial failure itself or pharmacological treatments of heart failure may cause alterations in cardiac structure, elevated CAR expression was not observed in heart failure entities other than DCM, in contrast with other alterations of myocardial structure and function associated with heart failure of various origins.19–23⇓⇓⇓⇓ Our findings may reflect a predisposing mechanism for cardiotropic viral infections in DCM hearts. However, CAR induction in cardiomyocytes has been reported in rat autoimmune myocarditis after the onset of infiltration by inflammatory cells.9 CAR upregulation triggered by nonviral factors may predispose for secondary viral infection, or primary viral infection may induce CAR upregulation. To establish such cause-effect relationships in humans, further studies on the relative time courses of CAR expression, virus load, and inflammation in human myocarditis are required.
Considering the possible pathogenic relevance of CAR regulation in humans, the absence of CAR from the luminal vascular surface and the low abundance in normal myocardium should render the normal human heart rather resistant to CAR-dependent viral infections. Although adenovirus infection is common, viral myocarditis due to adenoviruses is uncommon.24,25⇓ Because adenovirus myocarditis occurs both in children and adults,8 the question for “cardiac” predisposing factors facilitating adenovirus access to the human heart is of considerable interest. In cultured cardiomyocytes, we have shown directly that hCAR overexpression promotes very efficient adenovirus entry into these cells (Figure I in Data Supplement).
We propose that certain, as yet unknown factors (such as cytokines) upregulate CAR on human myocardium and that the level of this upregulation of the attachment receptor CAR26 is a major determinant of susceptibility to cardiac infection by adenoviruses and coxsackieviruses. The adenovirus coreceptors αvβ3 and αvβ527,28⇓ are present in equal amounts on both normal and diseased myocardium and are less likely to be key determinants. Because CAR is not only the attachment receptor for adenoviruses, but it also serves as a receptor for coxsackieviruses,1,2⇓ CAR may be similarly important for both of these otherwise unrelated viruses. Identification of the trigger factors that lead to CAR overexpression in the human heart is of considerable interest, and ablation of CAR-virus interaction3 may have therapeutic potential.
Expression Patterns and the Biological Role of CAR
Another hypothesis consistent with the data presented here concerns the normal biological role of CAR. Recombinant human CAR expressed in neonatal cardiomyocytes shows a very distinct preference for targeting to sites of cell-cell contact formation. Not only is CAR condensed at cell-cell boundaries between cardiomyocytes, but it is also at contact sites between cardiomyocytes and between cardiomyocytes and noncardiac cells. CAR may serve as a pathfinder protein expressed during embryogenesis to promote cardiac tissue formation that is then no longer needed to stabilize specialized cell-cell contacts. The hypothesis that CAR is a pathfinder protein explains the low CAR abundance in the adult, already well-structured heart as the expected normal state, whereas the observed CAR upregulation would be interpreted as part of an embryonic gene re-expression pattern generally seen in severe heart failure. This would not necessarily have adverse consequences in the case of CAR, but because of its nonphysiological role as a receptor for 2 viruses, the upregulation could have adverse consequences if a virus load is put on the patient. Given the low abundance of CAR in normal human myocardium, an essential role of CAR in maintaining the structural integrity of the adult myocardium seems unlikely, as opposed to a number of cytoskeletal, linkage, and extracellular proteins whose expression is dysregulated in failing human myocardium.18,27,28⇓⇓ However, upregulation of the putative pathfinder protein CAR may reflect an attempt of the organism at tissue repair by reestablishment of cell-cell contacts in a progressively dilating myocardium.
CAR Expression Patterns and Myocardial Gene Transfer
Absence of CAR from the human endothelial lumen may limit the access of intravascular virus to the myocardium and may be an important factor in limiting the tropism of CAR-dependent vectors in man. This is reminiscent of the situation in human airways, where restriction of CAR expression to the basolateral surfaces of the airway epithelium has been an obstacle to adenovirus-mediated gene therapy of cystic fibrosis.29–31⇓⇓ We found that CAR was expressed in subendothelial layers of the vessel wall; other investigators have reported the absence of CAR in the vessel wall of neonatal rats. These differences may reflect differences between tissues from neonatal rats and those from adult humans; alternatively, they may reflect differences in antibody specificity. Consequences of asymmetric CAR expression in the vascular wall for gene transfer need to be investigated further. Low basal CAR expression on the cardiomyocytes themselves may also limit adenoviral gene transfer, because the receptor complement on these target cells is incomplete. The influence of disease-associated or deliberately induced alterations of CAR expression on the efficiency and cellular targeting of myocardial gene transfer deserves attention.
This work was supported by the Deutsche Forschungsgemeinschaft through a Heisenberg Fellowship to Dr Poller (Po 378/2-1,2), grants from the Netherlands Heart Foundation to Dr de Jonge (97.221), and grants from the National Institutes of Health (HL54734), from Centeon Pharmaceutical Company, and from the Cardiovascular Research Center at Benjamin Franklin Hospital. We thank Kerstin Hinze for technical assistance (K. Hinze was supported by a grant from the Freie Universität Berlin).
The first 2 authors contributed equally to this work.
Additional information for the Methods section and an additional figure can be found in the Online Data Supplement at http://www.circulationaha.org
- ↵Bergelson J, Cunningham J, Droguett G, et al. Isolation of a common receptor for Coxsackie B viruses and adenoviruses 2 and 5. Science. 1997; 275: 1320–1323.
- ↵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.
- ↵Roelvink P, Lee G, Einfeld D, et al. Identification of a conserved receptor-binding site on the fiber proteins of CAR-recognizing adenoviridae. Science. 1999; 286: 1568–1571.
- ↵Bewley M, Springer K, Zhang Y-B, et al. Structural analysis of the mechanism of adenovirus binding to its human cellular receptor, CAR. Science. 1999; 286: 1579–1583.
- ↵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.
- ↵Pauschinger M, Dörner A, Kühl U, et al. Enteroviral RNA replication in the myocardium of patients with left ventricular dysfunction and clinically suspected myocarditis. Circulation. 1999; 99: 889–895.
- ↵Pauschinger M, Bowles N, Fuentes-Garcia J, et al. Detection of adenoviral genome in the myocardium of adult patients with idiopathic left ventricular dysfunction. Circulation. 1999; 99: 1348–1354.
- ↵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.
- ↵Eizema K, Fechner H, Bezstarosti K, et al. Adenovirus-based phospholamban-antisense-mRNA expression as a novel approach to improve cardiac contractile dysfunction: comparison of a constitutive viral versus an endothelin-1-responsive cardiac promoter. Circulation. 2000; 101: 2193–2199.
- ↵White D, Hata J, Shah A, et al. Preservation of myocardial β-adrenergic receptor signaling delays the development of heart failure after myocardial infarction. Proc Natl Acad Sci U S A. 2000; 97: 5428–5433.
- ↵Gambaryan S, Wagner C, Smolenski A, et al. cAMP-stimulated renin release from rat isolated perfused kidney, microdissected glomeruli, and isolated juxtaglomerular cells is inhibited by endogenous or overexpressed cGMP-dependent protein kinases. Proc Natl Acad Sci U S A. 1998; 95: 9003–9008.
- ↵Eizema K, vanHeugten H, Besztarosti K, et al. SERCA2 and ANF promotor-activity studies in hypertrophic cardiomyocytes using liposome-, gene gun- and adenovirus-mediated gene transfer.In: Takeda N, Nagano M, Dhalla N, eds. The Hypertrophied Heart. Vol 1. Boston: Kluwer Academic Publishers; 2000: 51–66.
- ↵Braunwald E, Bristow M. Congestive heart failure. Circulation. 2000; 102: IV-14–IV-23.
- ↵Horwitz M. Adenoviridae and their replication.In: Fields B, Knipe DM, Chanock RM, et al, eds. Virology. New York: Raven Press; 1990: 1679–1721.
- ↵Snavely S, Liu C, Adenovirus.In: Belshe R, ed. Textbook of Human Virology. Littleton, Mass: PSG Publishing Company; 1984: 779–794.
- ↵Roelvink P, Lizonova A, Lee J, et al. The coxsackievirus-adenovirus receptor protein can function as a cellular attachment protein for adenovirus serotypes from subgroups A, C, D, E, and F. J Virol. 1998; 72: 7909–7915.
- ↵Heling A, Zimmermann R, Kostin S, et al. Increased expression of cytoskeletal linkage and extracellular proteins in failing human myocardium. Circ Res. 2000; 86: 846–853.
- ↵Pickles R, McCarty D, Matsui H, et al. Limited entry of adenovirus vectors into well-differentiated airway epithelium is responsible for inefficient gene transfer. J Virol. 1998; 72: 6014–6023.