Molecular determinants of species specificity in the coronavirus receptor aminopeptidase N (CD13): influence of N-linked glycosylation.

Aminopeptidase N (APN), a 150-kDa metalloprotease also called CD13, serves as a receptor for serologically related coronaviruses of humans (human coronavirus 229E [HCoV-229E]), pigs, and cats. These virus-receptor interactions can be highly species specific; for example, the human coronavirus can use human APN (hAPN) but not porcine APN (pAPN) as its cellular receptor, and porcine coronaviruses can use pAPN but not hAPN. Substitution of pAPN amino acids 283 to 290 into hAPN for the corresponding amino acids 288 to 295 introduced an N-glycosylation sequon at amino acids 291 to 293 that blocked HCoV-229E receptor activity of hAPN. Substitution of two amino acids that inserted an N-glycosylation site at amino acid 291 also resulted in a mutant hAPN that lacked receptor activity because it failed to bind HCoV-229E. Single amino acid revertants that removed this sequon at amino acids 291 to 293 but had one or five pAPN amino acid substitution(s) in this region all regained HCoV-229E binding and receptor activities. To determine if other N-linked glycosylation differences between hAPN, feline APN (fAPN), and pAPN account for receptor specificity of pig and cat coronaviruses, a mutant hAPN protein that, like fAPN and pAPN, lacked a glycosylation sequon at 818 to 820 was studied. This sequon is within the region that determines receptor activity for porcine and feline coronaviruses. Mutant hAPN lacking the sequon at amino acids 818 to 820 maintained HCoV-229E receptor activity but did not gain receptor activity for porcine or feline coronaviruses. Thus, certain differences in glycosylation between coronavirus receptors from different species are critical determinants in the species specificity of infection.

Targeted disruption of the Ceacam1 (MHVR) gene leads to reduced susceptibility of mice to mouse hepatitis virus infection.

The CEACAM1 glycoproteins (formerly called biliary glycoproteins; BGP, C-CAM, CD66a, or MHVR) are members of the carcinoembryonic antigen family of cell adhesion molecules. In the mouse, splice variants of CEACAM1 have either two or four immunoglobulin (Ig) domains linked through a transmembrane domain to either a short or a long cytoplasmic tail. CEACAM1 has cell adhesion activity and acts as a signaling molecule, and long-tail isoforms inhibit the growth of colon and prostate tumor cells in rodents. CEACAM1 isoforms serve as receptors for several viral and bacterial pathogens, including the murine coronavirus mouse hepatitis virus (MHV) and Haemophilus influenzae, Neisseria gonorrhoeae, and Neisseria meningitidis in humans. To elucidate the mechanisms responsible for the many biological activities of CEACAM1, we modified the expression of the mouse Ceacam1 gene in vivo. Manipulation of the Ceacam1 gene in mouse embryonic stem cells that contained the Ceacam1a allele yielded a partial knockout. We obtained one line of mice in which the insert in the Ceacam1a gene had sustained a recombination event. This resulted in the markedly reduced expression of the two CEACAM1a isoforms with four Ig domains, whereas the expression of the two isoforms with two Ig domains was doubled relative to that in wild-type BALB/c (+/+) mice. Homozygous (p/p) Ceacam1a-targeted mice (Ceacam1aDelta4D) had no gross tissue abnormalities and were viable and fertile; however, they were more resistant to MHV A59 infection and death than normal (+/+) mice. Following intranasal inoculation with MHV A59, p/p mice developed markedly fewer and smaller lesions in the liver than +/+ or heterozygous (+/p) mice. The titers of virus produced in the livers were 50- to 100-fold lower in p/p mice than in +/p or +/+ mice. p/p mice survived a dose 100-fold higher than the lethal dose of virus for +/+ mice. +/p mice were intermediate between +/+ and p/p mice in susceptibility to liver damage, virus growth in liver, and susceptibility to killing by MHV. Ceacam1a-targeted mice provide a new model to study the effects of modulation of receptor expression on susceptibility to MHV infection in vivo.

Enteric infections with coronaviruses and toroviruses.

Many enteric viruses are difficult or impossible to propagate in tissue culture. Coronaviruses and toroviruses are large, enveloped, plus-strand RNA viruses in the order Nidovirales that cause enteric disease in young pigs, cows, dogs, mice, cats and horses. Two different serogroups of mammalian coronaviruses cause frequent respiratory infections in humans, and coronaviruses and toroviruses have been implicated in human diarrhoeal disease by immunoelectron microscopy. However, there is as yet no consensus about the importance of these enveloped viruses in human diarrhoea, and little is known about their genetic variability. The large spike (S) glycoprotein is an important determinant of species specificity, tissue tropism and virulence of coronavirus infection. To infect enterocytes, both S glycoproteins and the viral envelope must resist degradation by proteases, low and high pH, and bile salts. One specific site on the S glycoprotein of bovine coronavirus must be cleaved by an intracellular protease or trypsin to activate viral infectivity and cell fusion. S glycoprotein binds to specific receptors on the apical membranes of enterocytes, and can undergo a temperature-dependent, receptor-mediated conformational change that leads to fusion of the viral envelope with host membranes to initiate infection. Analysing spike-receptor interactions may lead to new ways to propagate these enteric viruses as well as new strategies for development of novel antiviral drugs.

Human coronavirus 229E infects polarized airway epithelia from the apical surface.

Gene transfer to differentiated airway epithelia with existing viral vectors is very inefficient when they are applied to the apical surface. This largely reflects the polarized distribution of receptors on the basolateral surface. To identify new receptor-ligand interactions that might be used to redirect vectors to the apical surface, we investigated the process of infection of airway epithelial cells by human coronavirus 229E (HCoV-229E), a common cause of respiratory tract infections. Using immunohistochemistry, we found the receptor for HCoV-229E (CD13 or aminopeptidase N) localized mainly to the apical surface of airway epithelia. When HCoV-229E was applied to the apical or basolateral surface of well-differentiated primary cultures of human airway epithelia, infection primarily occurred from the apical side. Similar results were noted when the virus was applied to cultured human tracheal explants. Newly synthesized virions were released mainly to the apical side. Thus, HCoV-229E preferentially infects human airway epithelia from the apical surface. The spike glycoprotein that mediates HCoV-229E binding and fusion to CD13 is a candidate for pseudotyping retroviral envelopes or modifying other viral vectors.

Comparison of expression patterns and cell adhesion properties of the mouse biliary glycoproteins Bbgp1 and Bbgp2.

Biliary glycoproteins are members of the carcinoembryonic antigen (CEA) family and behave as cell adhesion molecules. The mouse genome contains two very similar Bgp genes, Bgp1 and Bgp2, whereas the human and rat genomes contain only one BGP gene. A Bgp2 isoform was previously identified as an alternative receptor for the mouse coronavirus mouse hepatitis virus. This isoform consists of two extracellular immunoglobulin domains, a transmembrane domain and a cytoplasmic tail of five amino acids. In this report, we have examined whether the Bgp2 gene can express other isoforms in different mouse tissues. We found only one other isoform, which has a long cytoplasmic tail of 73 amino acids. The long cytodomain of the Bgp2 protein is highly similar to that of the Bgp1/4L isoform. The Bgp2 protein is expressed in low amounts in kidney and in a rectal carcinoma cell line. Antibodies specific to Bgp2 detected a 42-kDa protein, which is expressed at the cell surface of these samples. Bgp2 was found by immunocytochemistry in smooth muscle layers of the kidney, the uterus, in gut mononuclear cells and in the crypt epithelia of intestinal tissues. Transfection studies showed that, in contrast with Bgp1, the Bgp2 glycoprotein was not directly involved in intercellular adhesion. However, this protein is found in the proliferative compartment of the intestinal crypts and in cells involved in immune recognition. This suggests that the Bgp2 protein represents a distinctive member of the CEA family; its unusual expression patterns in mouse tissues and the unique functions it may be fulfilling may provide novel clues about the multiple functions mediated by a common BGP protein in humans and rats.

Neutralization of MHV-A59 by soluble recombinant receptor glycoproteins.

The interaction of viruses with specific receptors is an important determinant of viral tissue tropism and species specificity. Our goals are to understand how mouse hepatitis virus (MHV) recognizes its cellular receptor, MHVR, and how post-binding interactions with this receptor influence viral fusion and entry. Murine cells express a variety of cell surface molecule in the biliary glycoprotein (Bgp) family that are closely related to the MHVR. When these proteins are expressed at high levels in cell culture, they function as MHV receptors. We used a baculovirus expression system to produce soluble recombinant murine Bgp receptors in which the transmembrane and cytoplasmic domains have been replaced with a six-histidine tag. The soluble glycoproteins were purified to apparent homogeneity and shown to react with antisera to the native receptor. We compared the virus neutralizing activities of various soluble receptor glycoproteins. Soluble MHVR [sMHVR(1-4)] had 10-20 fold more virus neutralizing activity the soluble protein derived from the Bgp1b glycoprotein [sBgp1b(1-4)], from MHV-resistant SJL mice. The sMHVR(1-4) glycoprotein was 60-100 fold more active than a truncated receptor molecule containing only the first two immunoglobulin-like domains, sMHVR(1,2). The observation that sMHVR lacking domains 3 and 4 neutralizes MHV-A59 very poorly suggests that these domains may influence virus binding or subsequent steps associated with neutralization.

Virus-receptor interactions and interspecies transfer of a mouse hepatitis virus.

Molecular mechanisms regulating virus xenotropism and cross-species transmission are poorly understood. Host range mutants (MHV-H2) of mouse hepatitis virus (MHV) strains were isolated from mixed cultures containing progressively increasing concentrations of nonpermissive Syrian baby hamster kidney (BHK) cells and decreasing concentrations of permissive murine astrocytoma (DBT) cells. MHV-H2 was polytrophic, replicating efficiently in normally nonpermissive BHK cells, Syrian and Chinese hamster (DDT-1 and CHO) cells, human adenocarcinoma (HRT), primate kidney (VERO) and in murine 17Cl-1 cell lines. Little if any virus replication was detected in feline kidney (CRFK), and porcine testicular (ST) cell lines. To study the effects of xenotrophic spread on virus receptor-interactions in the original host, murine DBT cells were pretreated with a monoclonal antibody (MAb) CC1, directed against the MHV receptor, MHVR, a biliary glycoprotein (Bgp1a). Under treatment conditions that completely ablated the replication of the parental MHV strains, CC1 antireceptor antibodies did not block MHV-H2 replication. Following expression of MHVR in normally nonpermissive ST and CRFK cells, infection with the parental MHV strains, but not MHV-H2 was observed. To characterize the molecular basis preventing the interaction between MHV-H2 and MHVR, revertants of MHV-H2 (MHV-H2R6, MHV-H2R11) were isolated following a persistent MHV-H2 infection in DBT cells. These revertant viruses efficiently recognized MHVR, however infection of murine cells was resistant to MAb CC1 blockade. In addition, MHV-H2 and the revertant viruses efficiently recognized other Bgp receptors for docking and entry. These data suggest that interspecies transfer may remodel normal virus-receptor interactions that may result in altered virulence, tropism or pathogenesis in the original host.

Feline aminopeptidase N is a receptor for all group I coronaviruses.

Human coronavirus HCV-229E and porcine transmissible gastroenteritis virus (TGEV), both members of coronavirus group I, use aminopeptidase N (APN) as their cellular receptors. These viruses show marked species specificity in receptor utilization as they can only use APN of their respective species to initiate virus infection. Feline and canine coronaviruses are also group I coronaviruses. To determine whether feline APN could serve as a receptor for feline coronaviruses (FCoVs), we cloned the cDNA encoding feline APN (fAPN) by PCR from feline cells and stably expressed it in FCoV-resistant mouse or hamster cells. These became susceptible to infection with either of several biotypes of FCoVs. The expression of recombinant fAPN also made hamster and mouse cells susceptible to infection with other group I coronaviruses, including several canine coronavirus strains, transmissible gastroenteritis virus (TGEV), and human coronavirus HCV-229E. Thus, fAPN served as a functional receptor for each of these coronaviruses in group I. As expected, fAPN could not serve as a receptor for mouse hepatitis virus (MHV), a group II coronavirus which uses murine biliary glycoproteins as receptors. Thus, fAPN acts as a common receptor for coronaviruses in group I, in marked contrast to human and porcine APN glycoproteins which serve as receptors only for human and porcine coronaviruses, respectively. These observations suggest that cats could serve as a "mixing vessel" in which simultaneous infection with several group I coronaviruses could lead to recombination of viral genomes.

Role of mouse hepatitis virus-A59 receptor Bgp1a expression in virus-induced pathogenesis.

Expression of Bgp1a, a glycoprotein that serves as receptor for mouse hepatitis virus-A59 has been analyzed in various mouse tissues and correlated with the pathogenicity that this virus induces in the corresponding organs. Expression of Bgp1a was observed in many cells of epithelial origin, including hepatocytes and endothelial cells. It was also shown on macrophages and B lymphocytes. Bgp1a localization may easily explain infection and lysis of some cell types like hepatocytes. In contrast, other cell types that express the viral receptor are not infected after in vivo inoculation with mouse hepatitis virus-A59, which may be due to inaccessibility of the receptor to the virus during mouse infection, or to resistance to this virus in some cell types. This may account for the ability of the blood-brain barrier to prevent mouse hepatitis virus-A59 spreading into the central nervous system. In other organs, the virus may induce pathogenesis indirectly, resulting in the destruction of cells that do not express Bgp1a, like thymic lymphocytes, or else impair cell functions such as cytokine and immunoglobulin production by macrophages and B lymphocytes, respectively.

Selection in persistently infected murine cells of an MHV-A59 variant with extended host range.

Murine coronavirus MHV-A59 normally infects only murine cells in vitro and causes transmissible infection only in mice. In the 17 C1 1 line of murine cells, the receptor for MHV-A59 is MHVR, a biliary glycoprotein in the carcinoembryonic antigen (CEA) family of glycoproteins. We found that virus released from the 600th passage of 17 C1 1 cells persistently infected with MHV-A59 (MHV/pi600) replicated in hamster (BHK-21) cells. The virus was passaged and plaque-purified in BHK-21 cells, yielding the MHV/BHK strain. Because murine cells persistently infected with MHV-A59 express a markedly reduced level of MHVR (Sawicki, et al., 1995), we tested whether virus with altered receptor interactions was selected in the persistently infected culture. Infection of 17 C1 1 cells by MHV-A59 can be blocked by treating the cells with anti-MHVR MAb-CC1, while infection by MHV/BHK was only partially blocked by MAb-CC1. MHV/BHK virus was also more resistant than wild-type MHV-A59 to neutralization by purified, recombinant, soluble MHVR glycoprotein (sMHVR). Cells in the persistently infected culture may also express reduced levels of and have altered interactions with some of the Bgp-related glycoproteins that can serve as alternative receptors for MHV-A59. Unlike the parental MHV-A59 which only infects murine cells, MHV/BHK virus was able to infect cell lines derived from mice, hamsters, rats, cats, cows, monkeys and humans. However, MHV/BHK was not able to infect all mammalian species, because a pig (ST) cell line and a dog cell line (MDCK I) were not susceptible to infection. MHV/pi600 and MHV/BHK replicated in murine cells more slowly than MHV-A59 and formed smaller plaques. Thus, in the persistently infected murine cells which expressed a markedly reduced level of MHVR, virus variants were selected that have altered interactions with MHVR and an extended host range. In vivo, in mice infected with coronavirus, virus variants with altered receptor recognition and extended host range might be selected in tissues that have low levels of receptors. Depending upon the tissue in which such a virus variant was selected, it might be shed from the infected animal or eaten by a predator, thus presenting a possible means for initiating the transition of a variant virus into a new host as a model for an emerging virus disease.

Purified, soluble recombinant mouse hepatitis virus receptor, Bgp1(b), and Bgp2 murine coronavirus receptors differ in mouse hepatitis virus binding and neutralizing activities.

Mouse hepatitis virus receptor (MHVR) is a murine biliary glycoprotein (Bgp1(a)). Purified, soluble MHVR expressed from a recombinant vaccinia virus neutralized the infectivity of the A59 strain of mouse hepatitis virus (MHV-A59) in a concentration-dependent manner. Several anchored murine Bgps in addition to MHVR can also function as MHV-A59 receptors when expressed at high levels in nonmurine cells. To investigate the interactions of these alternative MHVR glycoproteins with MHV, we expressed and purified to apparent homogeneity the extracellular domains of several murine Bgps as soluble, six-histidine-tagged glycoproteins, using a baculovirus expression system. These include MHVR isoforms containing four or two extracellular domains and the corresponding Bgp1(b) glycoproteins from MHV-resistant SJL/J mice, as well as Bgp2 and truncation mutants of MHVR and Bgp1(b) comprised of the first two immunoglobulin-like domains. The soluble four-domain MHVR glycoprotein (sMHVR[1-4]) had fourfold more MHV-A59 neutralizing activity than the corresponding soluble Bgp1(b) (sBgp1(b)) glycoprotein and at least 1,000-fold more neutralizing activity than sBgp2. Although virus binds to the N-terminal domain (domain 1), soluble truncation mutants of MHVR and Bgp1(b) containing only domains 1 and 2 bound virus poorly and had 10- and 300-fold less MHV-A59 neutralizing activity than the corresponding four-domain glycoproteins. In contrast, the soluble MHVR glycoprotein containing domains 1 and 4 (sMHVR[1,4]) had as much neutralizing activity as the four-domain glycoprotein, sMHVR[1-4]. Thus, the virus neutralizing activity of MHVR domain 1 appears to be enhanced by domain 4. The sBgp1(b)[1-4] glycoprotein had 500-fold less neutralizing activity for MHV-JHM than for MHV-A59. Thus, MHV strains with differences in S-glycoprotein sequence, tissue tropism, and virulence can differ in the ability to utilize the various murine Bgps as receptors.

Mutational analysis of the virus and monoclonal antibody binding sites in MHVR, the cellular receptor of the murine coronavirus mouse hepatitis virus strain A59.

The primary cellular receptor for mouse hepatitis virus (MHV), a murine coronavirus, is MHVR (also referred to as Bgp1a or C-CAM), a transmembrane glycoprotein with four immunoglobulin-like domains in the murine biliary glycoprotein (Bgp) subfamily of the carcinoembryonic antigen (CEA) family. Other murine glycoproteins in the Bgp subfamily, including Bgp1b and Bgp2, also can serve as MHV receptors when transfected into MHV-resistant cells. Previous studies have shown that the 108-amino-acid N-terminal domain of MHVR is essential for virus receptor activity and is the binding site for monoclonal antibody (MAb) CC1, an antireceptor MAb that blocks MHV infection in vivo and in vitro. To further elucidate the regions of MHVR required for virus receptor activity and MAb CC1 binding, we constructed chimeras between MHVR and other members of the CEA family and tested them for MHV strain A59 (MHV-A59) receptor activity and MAb CC1 binding activity. In addition, we used site-directed mutagenesis to introduce selected amino acid changes into the N-terminal domains of MHVR and these chimeras and tested the abilities of these mutant glycoproteins to bind MAb CC1 and to function as MHV receptors. Several recombinant glycoproteins exhibited virus receptor activity but did not bind MAb CC1, indicating that the virus and MAb binding sites on the N-terminal domain of MHVR are not identical. Analysis of the recombinant glycoproteins showed that a short region of MHVR, between amino acids 34 and 52, is critical for MHV-A59 receptor activity. Additional regions of the N-terminal variable domain and the constant domains, however, greatly affected receptor activity. Thus, the molecular context in which the amino acids critical for MHV-A59 receptor activity are found profoundly influences the virus receptor activity of the glycoprotein.

Localization of mouse hepatitis virus open reading frame 1A derived proteins.

We have investigated the intracellular localization of proteolytic cleavage products encoded in the 5' portion of mouse hepatitis virus (MHV) gene 1. Immunofluorescent labeling of cells with an antiserum which recognizes p28, the ORF1a N-terminal cleavage product, resulted in widespread somewhat granular cytoplasmic staining, indicating that this protein is widely distributed in the cytoplasm of MHV-infected, but not control uninfected cells. Immunofluorescent staining of infected cells with antisera which recognize the downstream polypeptides, p65, p240 and p290 labeled discrete vesicular perinuclear structures. Double immunofluorescent labeling of BHK cells expressing the MHV receptor (BHK(MHVR1)) and infected with MHV-A59 with a Golgi-specific anti-mannosidase II monoclonal antibody and with antiserum recognizing each of these anti-MHV ORF1a polypeptides, showed that the p240 and p290 polypeptides were localized in discrete vesicular structures that overlapped the Golgi complex. Labeling with antibodies specific for p65 colocalized with the Golgi region, and showed staining of the perinuclear cytoplasm as well. Plasmids containing sequences contained in the first 6.75 kb of ORF1a have been expressed using the coupled vaccinia virus-T7 polymerase system. Immunofluorescent labeling of transfectants with the anti-ORF1a antisera showed patterns of antigen distribution similar to those observed in cells infected with MHV-A59. A deletion analysis with constructs containing only portions of the ORF1a sequence indicated that 303 amino acids containing the first papain-like protease domain (PLP-1) was sufficient to associate this protein with the Golgi.

The murine coronavirus mouse hepatitis virus strain A59 from persistently infected murine cells exhibits an extended host range.

In murine 17 Cl 1 cells persistently infected with murine coronavirus mouse hepatitis virus strain A59 (MHV-A59), expression of the virus receptor glycoprotein MHVR was markedly reduced (S. G. Sawicki, J. H. Lu, and K. V. Holmes, J. Virol. 69:5535-5543, 1995). Virus isolated from passage 600 of the persistently infected cells made smaller plaques on 17 Cl 1 cells than did MHV-A59. Unlike the parental MHV-A59, this variant virus also infected the BHK-21 (BHK) line of hamster cells. Virus plaque purified on BHK cells (MHV/BHK) grew more slowly in murine cells than did MHV-A59, and the rate of viral RNA synthesis was lower and the development of the viral nucleocapsid (N) protein was slower than those of MHV-A59. MHV/BHK was 100-fold more resistant to neutralization with the purified soluble recombinant MHV receptor glycoprotein (sMHVR) than was MHV-A59. Pretreatment of 17 Cl 1 cells with anti-MHVR monoclonal antibody CC1 protected the cells from infection with MHV-A59 but only partially protected them from infection with MHV/BHK. Thus, although MHV/BHK could still utilize MHVR as a receptor, its interactions with the receptor were significantly different from those of MHV-A59. To determine whether a hemagglutinin esterase (HE) glycoprotein that could bind the virions to 9-O-acetylated neuraminic acid moieties on the cell surface was expressed by MHV/BHK, an in situ esterase assay was used. No expression of HE activity was detected in 17 Cl 1 cells infected with MHV/BHK, suggesting that this virus, like MHV-A59, bound to cell membranes via its S glycoprotein. MHV/BHK was able to infect cell lines from many mammalian species, including murine (17 Cl 1), hamster (BHK), feline (Fcwf), bovine (MDBK), rat (RIE), monkey (Vero), and human (L132 and HeLa) cell lines. MHV/BHK could not infect dog kidney (MDCK I) or swine testis (ST) cell lines. Thus, in persistently infected murine cell lines that express very low levels of virus receptor MHVR and which also have and may express alternative virus receptors of lesser efficiency, there is a strong selective advantage for virus with altered interactions with receptor (D. S. Chen, M. Asanaka, F. S. Chen, J. E. Shively, and M. M. C. Lai, J. Virol. 71:1688-1691, 1997; D. S. Chen, M. Asanaka, K. Yokomori, F.-I. Wang, S. B. Hwang, H.-P. Li, and M. M. C. Lai, Proc. Natl. Acad. Sci. USA 92:12095-12099, 1995; P. Nedellec, G. S. Dveksler, E. Daniels, C. Turbide, B. Chow, A. A. Basile, K. V. Holmes, and N. Beauchemin, J. Virol. 68:4525-4537, 1994). Possibly, in coronavirus-infected animals, replication of the virus in tissues that express low levels of receptor might also select viruses with altered receptor recognition and extended host range.

Virus-receptor interactions in the enteric tract. Virus-receptor interactions.

Expression of specific virus receptors on the surface of intestinal epithelial cells or M cells can determine whether or not a animal is susceptible to infection with an enterotropic virus. Receptors for many animal viruses have been identified. The specificity of virus-receptor interactions clearly affects the species specificity of virus infection, and in some instances may be an important determinant of viral tissue tropism. In this paper, the specificity of coronavirus-receptor interactions is summarized. Porcine and human coronaviruses utilize aminopeptidase N as their receptors, but in a species-specific manner. Mouse hepatitis virus uses several rodent glycoproteins in the carcinoembryonic antigen family as receptors. In addition, some coronaviruses can interact with carbohydrate moieties on the cell surface. Understanding the molecular mechanisms of virus-receptor interactions may lead to development of novel strategies for the control of enteric viral diseases.

Role of virus receptor-bearing endothelial cells of the blood-brain barrier in preventing the spread of mouse hepatitis virus-A59 into the central nervous system.

BALB/c mice develop a neurologic demyelinating disease after inoculation of mouse hepatitis virus (MHV), strain A59, by the intracranial, but not by the intraperitoneal route. To determine the mechanisms that prevent virus spreading through the blood-brain barrier, we analyzed expression of MHVR, a glycoprotein that serves as receptor for mouse hepatitis virus on endothelial cells of cerebral blood vessels. Our results indicated that MHVR was strongly expressed on the endoluminal pole of these cells. In addition, a direct virus binding assay showed that mouse hepatitis virus was able to bind endothelial cells via this receptor. Despite this expression of a functional viral receptor, in normal mice infected with mouse hepatitis virus by the contra-peritoneal route, no in vivo viral replication could be detected in endothelial cells from the brain, contrasting with the equivalent cells from the liver. However, shortly after i.v. administration of sodium dodecylsulfate detergent to the mice, virus infection of some cerebral endothelial cells was detected in a few mice. As a consequence of detergent treatment, virus infection was able to cross the blood-brain barrier. These results suggest that the protective role of the blood-brain barrier against spreading of mouse hepatitis virus A59 into the central nervous system is determined by a specific restriction of viral entry into the endothelial cells of cerebral origin.