Expression of the Coxsackievirus and Adenovirus Receptor in Cultured Human Umbilical Vein Endothelial Cells: Regulation in Response to Cell Density

Abstract

Primary cultures of human umbilical vein endothelial cells (HUVEC) express the human coxsackievirus and adenovirus receptor (HCAR). Whereas HCAR expression in HeLa cells was constant with respect to cell density, HCAR expression in HUVEC increased with culture confluence. HCAR expression in HUVEC was not quantitatively altered by infection with coxsackievirus B.

The six serotypes of the group B coxsackieviruses (CVB1 to -6) are human enteroviruses that cause a variety of acute diseases and are strongly implicated as causative agents in several chronic diseases (1, 10). These enteroviruses presumably gain access to the heart, as well as to other tissues distant from the gut, via the blood during viremia. They must either pass through the vascular endothelium by transcytosis or infection or be carried past the endothelial barrier by infected circulating cells, which can migrate into the target tissues. In fact, two such events must occur, since the virus must first enter into the circulation and then leave the circulation to gain access to the secondary sites of infection. Both routes of dissemination are plausible, since both circulating cells and endothelial cells have been shown to be susceptible to CVB infection (6–8).

Some CVB strains can persistently infect endothelial cells in vitro, suggesting that endothelial cells may constitute a reservoir for long-term virus production and therefore be especially relevant to the development of chronic diseases (6). Preferential infection of endothelial cells from different tissues may be related to differential expression of the virus receptor (8), and it has been suggested that the persistent infection of endothelial cells may be due to down-regulation of the receptor in infected cultures (6). The general requirement for the human coxsackievirus-adenovirus receptor (HCAR) in mediating CVB3 infection (12) and the susceptibility of endothelial cells to CVB3 infection indicated that human umbilical vein endothelial cells (HUVEC) probably express the HCAR protein. We conducted the experiments reported here to document the expression of HCAR by HUVEC and to determine whether HCAR expression is constitutive or regulated in response to the culture environment.

Since HCAR appears to be the principal receptor mediating CVB infection, although infection may occur in some cases in the absence of HCAR (2, 11), we initially determined that HUVEC (Clonetics) express HCAR. Cell monolayers were rinsed with wash buffer (0.05 M Tris, 0.1 M NaCl [pH 7.5], and 0.02% sodium azide with 2 mM phenylmethylsulfonyl fluoride and 5 mM N-ethylmaleimide added from a 20× stock in ethanol) and scraped from the flask. An aliquot was taken for bicinchoninic acid protein assay (Pierce). Cells were then lysed with wash buffer containing 2% (wt/vol) octylglucoside. Proteins in the lysate were precipitated with acetone (five times the sample volume, on ice), and redissolved in sodium dodecyl sulfate-polyacrylamide gel electrophoresis sample solvent (9). Western blots of cell lysates were probed with monoclonal antibodies raised against the recombinant extracellular domain of HCAR expressed in Escherichia coli, followed by peroxidase-conjugated rabbit anti-mouse immunoglobulins (DAKO), and developed with ECL+Plus (Amersham) with BioMax film (Kodak). The single band detected in lysates of HeLa cells and HUVEC (Fig. 1A) corresponds to the HCAR protein identified on virus overlay blots (3), establishing that HUVEC express HCAR. To establish that differences in levels of expression of HCAR could be quantitated on Western blots, different amounts of HeLa cell lysates were applied to gels and bands on developed blots were integrated from images recorded with a flatbed scanner (Fig. 1B). The results show that HCAR expression is quantifiable by a Western blot-based method.

FIG. 1.

Analysis of HCAR expression by Western blotting. (A) Monoclonal antibodies raised against ECAR (the recombinant extracellular domain of HCAR) react specifically with the wild-type HCAR expressed by HeLa cells and HUVEC on Western blots. Octylglucoside lysates of HeLa cells and HUVEC were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis, blotted to Immobilon, and probed with monoclonal antibody E1. HCAR is indicated by >. Prestained protein standards were run in the center lane. RD cells, which are not typically infected by CVB, did not react with the antibody (blank line not shown). (B) Western blots developed by ECL+Plus provided quantitative responses for different amounts of HCAR applied to the gels. Predetermined amounts of total cell protein in HeLa cell lysates were analyzed on blots probed with monoclonal antibody E1. The data are compiled from 20 separate experiments. The inset shows typical results from a blot loaded with 25 to 150 μg of total HeLa cell protein in 25-μg increments from left to right.

To determine whether HCAR expression was altered in response to cell density, HUVEC and HeLa cells were plated in serial dilutions to provide a range of monolayer cell densities for analysis. HeLa cells were harvested the day after being subcultured, and HUVEC were analyzed within 24 to 48 h of being subcultured, with the medium being replenished the day before the experiment. HCAR expression in HeLa cells was unchanged over a wide range of cell densities (Fig. 2A), whereas HCAR expression in HUVEC was lowest in the least confluent cultures and greatest in the confluent cultures (Fig. 2B). As the ranges of protein concentrations and cell densities studied with HeLa cells encompassed those used in the analysis of HUVEC, it is unlikely that the differences between HeLa cells and HUVEC were due to disparate behaviors of the HCAR during the preparative procedures. These data demonstrate that HCAR expression in HUVEC differs from that in HeLa cells and appears to be regulated in response to HUVEC cell density.

FIG. 2.

HCAR expression in relationship to cell density in HUVEC is different from that in HeLa cells. (A) HCAR expression by HeLa cells is independent of cell density. The line was placed by linear regression analysis (including cultures containing up to 8,000 μg of protein/flask which are not plotted), and its slope was not significantly different from 0 (slope = −1.1 × 10⁻⁵, standard error of regression = 1.8 × 10⁻⁵). The inset shows the blot corresponding to the experiment whose results are plotted as open circles. (B) HCAR expression by HUVEC increases with the cell density of the culture. HUVEC at approximately 60% confluence (passage 5 or below) were trypsinized from several flasks, and serial dilutions of the lifted cells were seeded into new flasks. The subcultures were harvested and analyzed for protein content and HCAR expression after 24 to 48 h, when the flasks seeded at the highest cell densities had become confluent. All lanes were loaded with 150 μg of protein. The inset shows the blot corresponding to the experiment whose results are plotted as open circles. Cells per square centimeter shown on the ordinate axis were calculated from an analysis of cell numbers (determined by counting on a Coulter counter) paired with determinations of protein concentration.

Conaldi et al. (6) suggested that the ability of HUVEC to support chronic low-level infection by some CVB serotypes, including CVB3, might be the result of receptor down-regulation in response to the infection. To test this hypothesis, individual confluent HUVEC cultures were inoculated with CVB3/0 or CVB0 (4, 5, 13) on three consecutive days. Virus titers were then determined for the inoculated cultures. HCAR expressed by the infected HUVEC was quantitated and compared to HCAR expressed by control cultures that had not been inoculated. Virus titers increased more than 10-fold over background within 24 h and about 2-fold from 24 to 71 h postinfection (Fig. 3A), showing that the cell cultures harbored a productive, albeit low-level, infection throughout the course of the experiments. Other experiments (data not shown) demonstrated that infectious CVB3/0 was shed into the medium within 24 h of inoculation. CVB3/0 infection had no effect on HUVEC HCAR expression up to 3 days postinfection (Fig. 3A). Inoculation of HUVEC cultures with CVB3/0 (multiplicity of infection of 20), conditions which lyse HeLa cells within 24 h (Fig. 3B), had no obvious effects on HUVEC morphology (as assessed by light microscopy) or viability, up to 6 days postinoculation, consistent with results of a prior study (6). Although the relatively low level of infection and the lack of observed cytopathology might have been due to infection of a small subpopulation of HUVEC, it is evident that the persistence and lack of cytopathology were not due to a global down-regulation of HCAR in the cultures in response to either virus exposure or productive virus replication. Mechanisms which limit the CVB infection of HUVEC are apparently distinct from altered HCAR expression.

FIG. 3.

In contrast to HeLa cells neither culture density nor cell morphology were altered in HUVEC infected with CVB3/0. (A) Single-step growth curves (13) showed that HUVEC produced virus within 24 h of infection and continued to produce virus for at least 3 days (□). The amount of HCAR expressed was not altered by the infection (●, ▴). The inset shows the HCAR blot corresponding to the experiment whose results are plotted as triangles. (B) HeLa cells or HUVEC were seeded into 24-well plates and allowed to reach confluence. CVB3/0 was added to each experimental well, with uninfected wells serving as controls. At the indicated times following infection, the cells were fixed and stained with crystal violet.

This study demonstrates that HUVEC express HCAR, the principal cell surface receptor for coxsackieviruses and adenoviruses. Moreover, HUVEC appear to regulate HCAR expression as a function of cell density, in contrast to HeLa cells, which express essentially constant amounts of HCAR over a wide range of cell densities. Together, these results suggest that CVB have evolved to utilize a receptor maximally expressed on intact endothelium and that they may produce a chronic infection that provides better viral access to tissues underlying the endothelial layer.

HUVEC are derived from a unique, extraembryonic tissue, and limitations to extending the results of this study to issues of virus tropism and pathology in organs affected by coxsackievirus infections are acknowledged. Our results, however, are valid for this model and establish the precedent of HCAR regulation by endothelial cells. Future studies will determine whether similar regulation occurs in endothelial cells derived from coronary vessels as well as in other cell types derived from tissues involved in the pathology of coxsackievirus infection. Results from this and future studies not only will be pertinent to understanding the pathogenesis of these viruses but also may provide new information useful for manipulating potential targets of adenovirus-based gene therapy vectors.