Entry of Alphaherpesviruses Mediated by Poliovirus Receptor-Related Protein 1 and Poliovirus Receptor

See allHide authors and affiliations

Science  05 Jun 1998:
Vol. 280, Issue 5369, pp. 1618-1620
DOI: 10.1126/science.280.5369.1618


A human member of the immunoglobulin superfamily was shown to mediate entry of several alphaherpesviruses, including herpes simplex viruses (HSV) 1 and 2, porcine pseudorabies virus (PRV), and bovine herpesvirus 1 (BHV-1). This membrane glycoprotein is poliovirus receptor-related protein 1 (Prr1), designated here as HveC. Incubation of HSV-1 with a secreted form of HveC inhibited subsequent infection of a variety of cell lines, suggesting that HveC interacts directly with the virus. Poliovirus receptor (Pvr) itself mediated entry of PRV and BHV-1 but not of the HSV strains tested. HveC was expressed in human cells of epithelial and neuronal origin; it is the prime candidate for the coreceptor that allows both HSV-1 and HSV-2 to infect epithelial cells on mucosal surfaces and spread to cells of the nervous system.

Alphaherpesviruses, including HSV-1, HSV-2, PRV, and BHV-1, infect a variety of cell types in culture, resulting in efficient virus production in a short replicative cycle. Infection in the natural host is characterized by lesions in the epidermis, usually on mucosal surfaces, with spread of virus to the nervous system and establishment of latent infections in neurons. Binding of alphaherpesviruses to cells occurs primarily through an interaction of virion glycoprotein C (gC) with cell surface heparan sulfate, whereas fusion between the virion envelope and cell membrane requires the glycoproteins gB, gD, gH, and gL (1).

Several lines of evidence suggest that alphaherpesvirus gD interacts with a cell surface receptor in addition to heparan sulfate to mediate viral entry and that, in certain cell types, HSV-1, PRV, and BHV-1 can use a common gD receptor for entry (1, 2). Recently, a gD receptor for entry of HSV-1 and HSV-2 was identified as an additional member of the tumor necrosis factor receptor family, called herpesvirus entry mediator (3, 4), and is designated here as herpesvirus entry mediator A (HveA). HveA is the principal receptor for entry of HSV into human lymphoid cells but not into other cell types (3). Also, HveA fails to mediate the entry of PRV (3). A second mediator of HSV entry identified recently (5) was shown to be poliovirus receptor-related protein 2 (6). No function and no poliovirus receptor activity have been reported for this protein, and it is therefore designated herpesvirus entry mediator B (HveB). HveB mediates the entry of HSV-2 strains, PRV, and certain viable mutants of HSV-1 but fails to mediate the entry of wild-type HSV-1 strains or BHV-1 (5).

These results demonstrate that multiple alphaherpesvirus coreceptors exist, differing in their specificities for individual viruses in the subfamily. Neither HveA nor HveB fits the specifications for a coreceptor that can mediate entry of both HSV-1 and HSV-2 into epithelial cells at the initial site of infection and into neuronal cells for the establishment of latent infection. Also, neither HveA nor HveB serves as a coreceptor for all these viruses—HSV-1, PRV, and BHV-1—and therefore could not be a common coreceptor for these human and animal alphaherpesviruses. Because HveB is closely related to the poliovirus receptor (Pvr) (7) and to poliovirus receptor-related protein 1 (Prr1) (8), we explored the possibility that one or both of those proteins might mediate the entry of HSV-1 and -2 as well as PRV and BHV-1.

Chinese hamster ovary (CHO) cells express heparan sulfate chains, to which alphaherpesviruses can bind. However, CHO cells are resistant to the entry of HSV-1, PRV, and BHV-1 because of the absence of coreceptors required for virion cell fusion (3, 5, 9). CHO cells were transfected with plasmids expressing Pvr or Prr1 (10) and then inoculated with virus (11) to determine whether expression of the cell proteins could provide the necessary coreceptors for viral entry. Prr1 mediated the entry of several HSV-1 strains and three HSV-1 mutants (ANG, Rid1, Rid2) with amino acid substitutions in gD that preclude the use of HveA for entry (3) (Fig. 1). Prr1, designated here as HveC, also enhanced infection by HSV-2 strains (Fig. 1), although the enhancement was not as great because control CHO cells are partially susceptible to HSV-2 infection (9). HveC expression rendered CHO cells susceptible to PRV and BHV-1 as well as HSV, entry being a function of virus dose (Fig.2). Three independently isolated cell lines (CHO–HveC-1, -2, and -3) stably expressing HveC were also capable of being infected by PRV, BHV-1, and HSV-1 (12). Pvr mediated the entry of PRV and BHV-1 but not of the HSV-1 strains (Fig.2). The fact that HveC and Pvr, designated Pvr-HveD, can mediate entry of PRV and BHV-1 does not strictly imply that human cells could support the replication of those viruses; it does suggest, however, that the animal homologs of HveC and Pvr-HveD could mediate entry of those viruses into cells of the natural hosts.

Figure 1

Enhanced entry of HSV-1 and -2 strains into CHO cells expressing HveC. Subconfluent CHO-IEβ8 cells were transfected with a plasmid expressing HveC (pBG38, solid bars) or a control plasmid (pcDNA3, open bars) and 24 hours later replated in 96-well plates (about 2 × 104 to 4 × 104 cells per well). After 24 hours, the cells were incubated with virus at a range of concentrations and β-Gal activity was quantitated as a measure of viral entry, as described (3,11). The results depicted were for 50,000 pfu per well, in the linear range of plots of virus dose against β-Gal activity. Each experiment was performed with a subset of the viruses that always included HSV-1(KOS). Within each experiment, all values were made relative to the value obtained for the HSV-1(KOS)/HveC infection. Each virus was inoculated in triplicate; the mean values plus SDs for at least two separate experiments are shown.

Figure 2

Enhanced entry of HSV-1, PRV, and BHV-1 into HveC- and Pvr-HveD–expressing CHO cells. Subconfluent CHO-IEβ8 cells (3, 11) were transfected with plasmids expressing HveA (pBec10) (3), HveC (pBG38), Pvr-HveD (pBG42.16), or control DNA (pcDNA3). After 24 hours, the transfected cells were replated in 96-well plates (about 2 × 104 to 4 × 104 cells per well) and exposed the next day to HSV-1(KOS), HSV-1(KOS)Rid1, PRV(Kaplan), or BHV-1(Cooper) isolates expressing β-Gal (11). Six hours after inoculation, cells were lysed and β-Gal activity was determined as a measure of virus entry (11). The infections were done in triplicate and were repeated four times. The mean values plus SDs for a representative experiment are shown.

The nucleotide sequence of our isolate of HveC cDNA (GenBank accession number AF060231), as well as I.M.A.G.E. Consortium Clone 287663 (a fragment of HveC cDNA isolated from brain tissue) (13) differed from the originally published sequence (8) by the absence of single bases at positions 582, 597, and 617. The differences affected the amino acid sequence over a short range but maintained the overall open reading frame (14). Diagnostic restriction enzyme analysis of our clone, as well as the corresponding region amplified by polymerase chain reaction (PCR) from a HeLa cell cDNA library, verified our sequence (12). The hveC gene is located on human chromosome 11 (8), which is of special interest because a human gene capable of conferring to Chinese hamster lung cells a susceptibility to HSV-1 infection was previously mapped to human chromosome 11 (15). The genes for HveB and Pvr-HveD are located on chromosome 19 (6, 16), whereas that for HveA is on chromosome 1 (17).

To identify human cell types in which entry of HSV-1 might be mediated by HveC, we performed reverse transcription–PCR (RT-PCR) analysis with primers specific for HveC cDNA on total RNA isolated from human cell lines and primary cells (18). HveC mRNA expression was detected in NT2 cells (teratocarcinoma), SH-SY5Y and IMR-5 cells (neuroblastomas), HL-60 cells (promyelocytic leukemia), primary human diploid fibroblasts, primary human foreskin keratinocytes (Fig.3), and HeLa cells (12) but not in HEL299 cells (embryonic lung fibroblasts) or phytohemagglutinin-activated T cell blasts (Fig. 3) (18). As expected, expression of HveC mRNA was also detected in CHO cells that were stably expressing HveC cDNA but not in control CHO cells or in CHO cells stably expressing HveA or HveB (Fig. 3). RT-PCR performed with primers specific for HveA cDNA yielded the expected product in keratinocytes and T lymphoblasts but not in NT2, SH-SY5Y, or IMR-5 cells (12). NT2, SH-SY5Y, and IMR-5 cells are susceptible to HSV-1(KOS) infection (12) (Fig.4). HveC is the best candidate for the entry protein used by HSV-1(KOS) in these cells because expression of HveC, but not of HveA, was detected and neither HveB nor Pvr-HveD mediates HSV-1(KOS) entry. Although all four of the herpesvirus entry proteins are expressed in many human tissues and organs, expression of HveA is detected principally in lymphoid organs (17, 19), whereas HveB, HveC, and Pvr-HveD can be expressed in cells of the nervous system (5, 7) or in cells cultured from the nervous system (Fig. 3).

Figure 3

Expression of HveC mRNA. Total RNA was isolated from established cell lines and primary cells, the cDNA was obtained, and PCR was performed as described (18). (A) HveC mRNA expression in CHO lines stably expressing HveA (HveA-12), HveB (HveB-1), or HveC (HveC-1) and also the parental CHO cell line (K1) (18). (B) HveC mRNA expression in human cell lines as described in the text (18) and in primary cell cultures: HDF, human diploid fibroblasts; HuFK, human foreskin keratinocytes; HuTL, phytohemagglutinin-activated human T cell blasts (18). (C) β-Actin mRNA expression detected in the RNA samples from the corresponding lanes in (B) (18). The β-actin control results are also presented in a figure demonstrating HveB expression in these same cell lines (5).

Figure 4

Blocking HSV-1 infection by secreted HveA and HveC. HSV-1(KOS)tk12 was preincubated with BSA, HveA(200t), or HveC(346t) for 1 hour at 37°C. The cell lines (about 4 × 104 cells per well) were then exposed to the mixtures of virus and protein in 96-well plates for 1 hour at 4°C before transfer to 37°C for 6 hours. CHO–IEβ8/HveA (CHO-HveA) cells and CHO–IEβ8/HveC (CHO-HveC) cells were generated by transfection of CHO–IEβ8 cells (3) as described (21). Cells were lysed and β-Gal activity was quantitated as a measure of infection by HSV-1(KOS)tk12 (20). The values obtained for infections in the presence of added protein were compared with the infection obtained in the absence of added protein to determine percent of control. All values represent the average of at least two experiments performed in triplicate.

A secreted form of HveA, HveA(200t), binds to virus by an interaction with HSV-1 gD and blocks infection by HSV-1 (4). To ascertain whether HveC interacts with virion proteins to mediate entry, we incubated HSV-1 with HveA(200t) or with a secreted form of HveC, HveC(346t) (20), prior to infection. Incubation of HSV-1 with either of these truncated proteins inhibited infection of NT2 cells, IMR-5 cells, CHO cell lines stably expressing HveA (21) or HveC (22) (Fig. 4), and SH-SY5Y cells (23) in a dose-dependent manner. Either protein competed with membrane-bound forms of HveA or HveC for interaction with virus, thereby inhibiting infection. As expected, HveA and HveC were expressed on the surface of CHO–HveA-12 and CHO–HveC-1 cells, respectively, according to flow cytometric analysis (3, 12). Recent results demonstrate that HveC(346t) binds purified gD from HSV-1 and HSV-2 in vitro (22). Therefore, HveC(346t) may bind the gD present on the HSV-1 virion to inhibit infection, and HveC may be a receptor for virion gD.

Whereas HveA, HveB, and Pvr-HveD are active for entry of subsets of the alphaherpesviruses tested here, HveC mediated entry of all these viruses. Although the gD family members of the alphaherpesviruses share only 10 to 15% amino acid sequence identity, there is general conservation of the positions of six cysteine residues and probably conservation of a domain recognized by HveC. Substitutions at position 27 in HSV-1 gD abrogate entry via HveA (3) and enable entry via HveB (5) but have no effect on entry via HveC. These findings, coupled with the ability of HveA and HveC to compete for critical sites on virions to block infection, indicate that each entry protein may recognize overlapping but distinct structural domains of gD. Animal homologs of HveB, HveC, and Pvr-HveD probably are the principal coreceptors for entry of animal alphaherpesviruses and may be active for HSV strains, as predicted by experiments that identified gD coreceptors recognized by both animal and human alphaherpesviruses (2); however, this remains to be determined.

HSV-1 and HSV-2 strains exhibit differences in pathogenesis, some of which may be attributable to preferences for different entry receptors and, therefore, to targeting of different cell types. However, a common feature of all HSV-1 and HSV-2 strains, and of most alphaherpesviruses, is the ability to replicate in mucosal epithelia and to invade adjacent nerve endings, thereby establishing a latent infection in nerve cell bodies. The results presented here implicate HveC as the prime receptor allowing for HSV-1 and HSV-2 infection of mucosal surfaces and spread to the nervous system and suggest it as a prime target for innovative prophylactic or therapeutic interventions. Moreover, HveC homologs may serve a similar role for infections by animal alphaherpesviruses of their natural hosts and thus may account for the cross-interference patterns observed for human and animal alphaherpesviruses.

  • * To whom correspondence should be addressed. E-mail: p-spear{at}


Stay Connected to Science

Navigate This Article