Journal of General Virology

Mutational analysis of the major heparan sulfate-binding domain of herpes simplex virus type 1 glycoprotein C

Kristina Mårdberg,¹ Edward Trybala,¹ Joseph C. Glorioso² and Tomas Bergström¹

¹ Department of Virology, Göteborg University, Guldhedsgatan 10b, S-413 46 Göteborg, Sweden
² Department of Molecular Genetics and Biochemistry, University of Pittsburgh School of Medicine, Pittsburgh, PA 15261, USA

Heparan sulfate (HS) has been identified as a receptor molecule for numerous microbial pathogens, including herpes simplex virus type 1 (HSV-1). To further define the major HS-binding domain of the HSV-1 attachment protein, i.e. glycoprotein C (gC), virus mutants carrying alterations of either two neighbouring basic amino acid residues or a single hydrophobic amino acid residue within the N-terminal domain of the protein (residues 26–227) were constructed. In addition, a mutant lacking the Asn148 glycosylation site was included in the study. Binding of purified mutated gC proteins to isolated HS chains showed that viruses with mutations at residues Arg(129,130), Ile142, Arg(143,145), Arg(145,147), Arg(151,155) and Arg(155,160) had significantly impaired HS binding, in contrast to the other mutations, including Asn148. Impairment of the HS-binding activity of gC by these mutations had profound consequences for virus attachment and infection of cells in which amounts of HS exposed on the cell surface had been reduced. It is suggested that basic and hydrophobic residues localized at the Cys127–Cys144 loop of HSV-1 gC constitute a major HS-binding domain, with the most active amino acids situated near the C-terminal region of the two cysteines.

Cell surface heparan sulfate (HS), a structurally diverse polysulfated glycosaminoglycan (Gallagher, 1995; Salmivirta et al., 1996), was first reported to serve as an initial receptor for herpes simplex virus (HSV) (WuDunn & Spear, 1989). Interaction with this molecule has thereafter been found to be a common pathway for attachment that is utilized by several human and animal viruses. Of the Herpesviridae family, all mammalian herpesviruses from the subfamily -Herpesvirinae, i.e. HSV types 1 and 2, varicella-zoster virus, pseudorabies virus (PrV) and bovine herpesvirus type 1 (BHV-1) as well as cell culture-adapted strains of equine herpesvirus type 1 were found to bind to HS (WuDunn & Spear, 1989; Mettenleiter et al., 1990; Okazaki et al., 1991; Zhu et al., 1995).

At least two HSV envelope glycoproteins, glycoprotein B (gB) and C (gC), may mediate virus binding to cell surface HS (Herold et al., 1991, 1994; Gerber et al., 1995; Laquerre et al., 1998). The finding that HSV-1 mutants deficient in the expression of both gB and gC (Herold et al., 1994) as well as gC-null (gC^–) constructs lacking a putative heparin-binding domain of gB (Laquerre et al., 1998) were severely impaired in HS-dependent attachment indicated that other envelope glycoproteins are less important during the virus attachment phase. A dominant role for gC over gB during virus attachment was suggested by the ability of some gC- but not gB-reactive antibodies to block virus binding (Fuller & Spear, 1985; Svennerholm et al., 1991). Furthermore, gC^– mutants showed, in contrast to gB^– mutants, an impairment in binding to some cell cultures (Homa et al., 1986; Herold et al., 1991, 1994; Tal-Singer et al., 1995). Hence, gC could be regarded as the principal attachment protein, acting through its HS-binding ability, although gB provides functional redundancy that might be important during infection of some cell types such as neurons (Immergluck et al., 1998).

HSV-1 gC contains, as do many heparin-binding proteins, several clusters of basic residues that enable an electrostatic interaction with the negatively charged sulfate and carboxylate groups on the HS molecule. In this report, we have characterized a major HS-binding domain by performing site-directed mutagenesis on all of the positively charged residues and some of the hydrophobic residues in the N-terminal domain of gC. Results from binding assays of purified gC with different mutations to isolated HS chains showed that a cluster of arginine residues surrounding the Cys127–Cys144 disulphide bridge (Rux et al., 1996) were required for this interaction and that Ile142 substantially contributed to the binding. Furthermore, in cells that were enzymatically pre-treated in order to reduce the number of exposed HS chains, the same HSV-1 gC mutants which were impaired in HS-binding also showed reduced attachment and infectivity.

Cells, viruses and antibodies. African green monkey kidney (GMK-AH1), human epidermoid carcinoma-2 (HEp-2) and Vero cells were grown in Eagle's minimum essential medium (EMEM) supplemented with 4 % heat-inactivated foetal bovine serum (FBS) and 0.05 % primaton RL substance (GMK-AH1 cells), 8 % FBS (HEp-2 cells) or 5 % heat-inactivated FBS and 1 % tricine (Vero cells). The HSV-1 strains used in this study were KOS321 and its gC^– derivative, designated gC^–39 (Holland et al., 1984). The preparation of virus stocks and determination of their titres were carried out in GMK-AH1 cells. The gC-reactive monoclonal antibodies (MAbs) that were utilized in this study were C1, C2 and C11, previously mapped to antigenic site I, and C3, C8, C10 and C13, which react with antigenic site II (Marlin et al., 1985; Wu et al., 1990). In addition, MAb B1C1, which also binds to antigenic site II, and MAbs C4H11B6 and C2H12 as well as the rabbit antiserum KF922 were used (Bergström et al., 1992; Trybala et al., 1994; Olofsson et al., 1999).

DNA purification. gC^–39 DNA, used for co-transfection, was purified from extracellular virions grown in GMK-AH1 cells. Briefly, cells cultured in 1 litre roller bottles were infected with virus at an m.o.i. of 1 p.f.u. per cell and incubated at 37 °C for 2 h. Cells were then rinsed once with EMEM and incubated for a further 48 h in 40 ml of fresh EMEM. Medium was collected and clarified by centrifuging at 2000 r.p.m. for 10 min. Virus was pelleted from the supernatant by centrifugation at 140000 g for 1 h, resuspended in 2 ml of Tris–EDTA (TE) buffer containing 10 mM Tris–HCl and 1 mM EDTA, pH 7.5, and disrupted with a solution containing 0.5 % SDS, 10 mM EDTA and 0.2 M NaCl. Viral DNA was then purified by phenol–chloroform extraction and precipitated with ethanol. DNA from HSV-1 gC mutants, used for sequencing and Southern blot analysis, was obtained from extracellular virions grown in plastic bottle cultures of GMK-AH1 cells. Virus was pelleted by ultracentrifugation and the pellet was resuspended in 400 µl of TE buffer. DNA was then purified using the QIAamp Blood kit (Qiagen). All restriction enzyme digestions and ligations of plasmid DNA were performed following standard recombinant DNA techniques (Maniatis et al., 1982).

Construction and identification of HSV gC mutants. The pGC plasmid comprising the gC gene (Homa et al., 1986) was cleaved with BsaAI (at nucleotide 605) and NheI (at nucleotide –34). The 639 bp fragment encoding the major part of antigenic site II of gC was inserted into the polylinker segment of pALTER between the SmaI and XbaI restriction enzyme sites. The resulting plasmid, pAltgC, served as a matrix for further mutations. Site-directed mutagenesis was performed using the Altered Sites II in vitro Mutagenesis system (Promega). The template plasmid, pAltgC, was rendered single-stranded, according to the protocol of the Altered Sites system, using the helper phage R408. To enhance mutagenesis, the selection primer was used at a 1:10 and a 1:500 dilution for each mutant single-stranded polynucleotide. Mutagenesis primers were designed to convert two neighbouring positively charged amino acids (arginine or lysine) to alanine or to introduce a single amino acid mutation (Fig. 1). For detection purposes, each primer also carried a new restriction enzyme site. Once mutagenesis was confirmed by enzyme analysis, pAltgC was digested with BglII and EcoNI (at nucleotides –18 and 525, respectively, of the gC gene). The resulting 543 bp BglII–EcoNI fragment was inserted into its wild-type position in pGC, i.e. between the BglII and EcoNI sites, and sequenced to determine the presence of the desired mutations and also the lack of any unwanted mutations. DNA sequencing was carried out using dye-terminator cycle sequencing and an ABI Prism 310 Genetic Analyser (Perkin Elmer).

Fig. 1. Specific amino acid residues of HSV-1 gC, indicated by numbers and the one letter amino acid code, were altered by site-directed mutagenesis to determine their role in binding to HS. All positively charged amino acids (indicated in red) between residues 89 and 160 were mutated in pairs to alanine, a non-charged residue. Also, three mutants that carry a single amino acid substitution were constructed. The two hydrophobic residues V140 and I142 (indicated in blue) were replaced by the non-hydrophobic amino acid threonine, and N148 was mutated to alanine, thereby abolishing an N-glycosylation site (indicated in green). Filled circles indicate residues decisive for the HS-binding capacity of gC.

Subconfluent Vero cells (60–80 %) were co-transfected with a mixture of 2 µg of gC^–39 DNA, 2 µg of SalI-digested wild-type or point-mutated pGC and 10 µg of calf thymus DNA using the calcium–phosphate co-precipitation method (Graham & van der Eb, 1973). After 4 h of incubation, cells were shocked with 25 % glycerol for 4 min, rinsed and incubated in EMEM supplemented with 10 % foetal calf serum until cytopathic effect (CPE) was widespread (usually 4 to 5 days). Progeny virus was then plaque-purified on monolayers of Vero cells using the immunoreactive black plaque assay (Holland et al., 1983). A mixture of anti-gC MAbs (C4H11B6, C2H12 and B1C1) was used for staining purposes and the purification procedure was regarded as complete when 100 % of the plaques were stained in two successive rounds of purification. To confirm the genotype of mutant viruses, viral DNA was sequenced and Southern blot was performed using the DIG system (Boehringer Mannheim). Hybridizations were carried out overnight at 69 °C in DIG Easy Hybridization solution using a 288 bp probe (synthesized by PCR) with incorporated DIG-labelled nucleotides complementary to the promoter signal sequence region of gC.

Reactivity of HSV-1 gC mutants with MAbs. The reactivity of MAbs with HSV-1 gC mutant viruses was tested by an ELISA-based method, as described previously (Trybala et al., 1994). Briefly, duplicate monolayers of GMK-AH1 cells in 96-well plates were infected with the different virus constructs at an m.o.i. of 50 p.f.u. per cell. HSV-1 strain KOS321 was used as a positive control and gC^–39 as a negative control. When CPE was complete, monolayers were fixed with 0.25 % glutaraldehyde, and 100 µl EMEM containing either of the gC-specific MAbs was then added. Following incubation of the cells with alkaline phosphatase-conjugated goat anti-mouse IgG (Jackson ImmunoResearch Laboratory) and, subsequently, substrate, the absorbance value of the supernatant was measured at 405 nm.

To determine whether the introduced mutations in gC affected the expression of gC in mutant virions, we used an ELISA in which purified virus particles from the different mutants, previously quantified by their DNA content, were used as antigens. In brief, 20 µg of each mutant virus, serially diluted in carbonate buffer (eight times in twofold steps), was added to 96-well plates in quadruplicate and left overnight at 4 °C. Thereafter, 100 µl PBS with 1 % BSA containing either a gC-reactive MAb (C4H11B6), of which the affinity towards gC was unaltered by the mutations introduced in gC, or the gE-reactive MAb B1E6, was added in duplicate. Conjugate was added and the absorbance values were measured at 405 nm, as described above. Quantification of gC was performed by comparing the reactivity of MAb C4H11B6 at the different concentrations of virus constructs. In addition, a ratio of the virus concentrations giving identical absorbance values within linear parts of the curves for the anti-gC and anti-gE MAbs was calculated for each mutant virus.

Purification of HSV-1 gC. GMK-AH1 cells in 1 litre roller bottles were infected with KOS321, gC^–39 or HSV-1 gC mutants at an m.o.i. of 1 p.f.u. per cell and incubated at 37 °C for 36 h. The infected cells were centrifuged at low speed and disrupted by freezing. The extracellular virus was pelleted from the supernatant medium by centrifugation at 160000 g for 1 h. Virus and cell pellets were combined, resuspended in cold lysis buffer [0.02 M Tris–HCl, 0.15 M NaCl, 1 % Nonidet P-40, 1 % sodium deoxycholate, 1 mM EDTA and 2 mM 4-(2-aminoethyl)-benzenesulfonyl fluoride, pH 7.5] and kept on ice for 1 h. The lysate was then clarified by centrifugation at 130000 g for 1 h. Since some gE contamination in our immunosorbent purified gC preparations had been noticed previously, lysates were first passed through an immunosorbent column containing the gE-specific MAb B1E6 (Bergström et al., 1992) and then run through a column with MAb C4H11B6 to adsorb gC. Subsequently the column was washed with 0.02 M Tris–HCl buffer (pH 7.5) containing 0.5 M NaCl, 0.1 % Nonidet P-40 and 1 mM EDTA and then again with the same buffer without detergent. Bound material was eluted with 0.1 M glycine–HCl buffer (pH 2.4) and immediately neutralized with 1 M Tris–HCl, pH 8.0. The fractions that contained the highest amounts of gC were pooled and centrifuged through 30 kDa filters (Palfiltron) to exchange the eluent buffer with PBS (137 mM NaCl, 2.7 mM KCl, 8.1 mM Na₂HPO₄ and 1.5 mM KH₂PO₄). The protein concentration in the purified gC preparations was determined by the standard Lowry method (BioRad). The differently mutated gC proteins were then subjected to electrophoresis in a preformed separation gel of 4–12 % polyacrylamide (Novex) and stained with colloidal Coomassie blue.

Isolation of radiolabelled HS chains and binding of gC to HS. To radiolabel glycosaminoglycans, subconfluent monolayers of HEp-2 and GMK-AH1 cells were grown for 48 h in the presence of Na₂³⁵SO₄ (50 µCi/ml, Amersham Life Sciences) in sulfate-free EMEM supplemented with antibiotics and 8 % FBS for HEp-2 cells, and with 0.05 % primaton RL substance and 2 % newborn calf serum for GMK-AH1 cells. Cell-associated HS chains were isolated as described previously (Lyon et al., 1994). For the HS-binding assay, purified gC (0.5 µg) in 0.2 ml of PBS supplemented with 0.05 % BSA was mixed with approximately 3000 c.p.m. of ³⁵S-labelled HS chains isolated from HEp-2 and GMK-AH1 cells. The rest of the procedure was carried out as described previously (Trybala et al., 1998).

Haemagglutinating activity (HA) of HSV-1 gC mutants. HSV-1 can induce a gC-dependent, heparinase-sensitive HA of mouse erythrocytes and an assay to determine the HA of HSV-1 gC was carried out as described previously (Trybala et al., 1993). In brief, cell-associated (CA) and extracellular (EX) HA antigens were prepared from virus-infected GMK-AH1 cells and infectious medium, respectively. Erythrocytes were collected from C57BL/6 mice, approximately 2 months old, and washed with an isotonic solution of NaCl buffered with 0.02 M phosphates (Na₂HPO₄ and KH₂PO₄, pH 7.0). To determine HA, serial dilutions of EX and CA preparations were mixed with murine erythrocytes and results expressed as the highest virus dilution that induced haemagglutination.

Purification of radiolabelled EX virus. GMK-AH1 cells in roller bottles were infected with mutant viruses, KOS321 or gC^–39 at an m.o.i. of 3 p.f.u. per cell. After virus adsorption for 3 h at 37 °C, cells were washed twice with EMEM before the addition of 45–50 ml of EMEM supplemented with [methyl-³H]thymidine (40 µCi/ml). Cells were incubated for 48 h and the culture medium was collected and centrifuged at 2500 r.p.m. for 25 min to remove cell debris. The supernatant was then centrifuged at 36000 g for 2 h and the virus pellet was covered with 0.2 ml of PBS and left at 4 °C overnight. The pellet was gently resuspended in PBS and purified on a three-step discontinuous gradient of sucrose, as described previously (Karger et al., 1995). Unless otherwise stated, purified virus was resuspended in PBS containing 0.1 % BSA and stored at –70 °C. The number of virus particles in the purified preparations was calculated based on the determination of the DNA content (Karger et al., 1995). The expression of gC by mutated HSV-1 virions was confirmed by immunoblotting using purified virus particles as antigen and the polyclonal anti-gC rabbit serum KF922.

Binding of radiolabelled virus to cells pre-treated with heparinase. The following experiment was performed to determine the degree of HS reduction (cleavage) on cell surfaces allowing optimal discrimination between gC-positive (KOS321) and gC-negative virus (gC^–39) attachment. Duplicate monolayers of HEp-2 cells were pre-treated with serial twofold dilutions of heparinase ranging from 8 to 0.008 units/ml for 1 h at 37 °C. The cells were then cooled for 30 min at 4 °C, washed twice with cold PBS and blocked for 1 h at 4 °C with PBS supplemented with 1 % BSA. The respective virus strains were added at identical titres (25000 virus particles per cell) and left for adsorption for 1 h at 4 °C with continuous shaking. Cells were then washed three times in cold PBS to remove unadsorbed virions and lysed in 5 % SDS. Radioactivity was determined by scintillation and results were expressed as a percentage of the number of virus particles that attached to the mock-treated control cells. Binding of HSV-1 gC mutants to heparinase-treated cells was assayed in a similar manner except that duplicate monolayers of GMK-AH1 and HEp-2 cells were pre-treated with a selected concentration of heparinase (0.25 units/ml).

Infectivity of HSV-1 gC mutants on heparinase-treated cells. Confluent monolayers of GMK-AH1 cells in 6-well plates were rinsed twice with 2 ml of EMEM and then treated for 1 h at 37 °C with 1 unit of heparinase resuspended in 1 ml EMEM containing 0.1 % BSA. The cells were again rinsed and approximately 100 p.f.u. of the respective mutant virus in 1 ml of EMEM was added and left to adsorb for 30 min at 37 °C. Cells were then washed twice with EMEM and overlaid with 4 ml of 1 % methylcellulose solution. After 3 days of incubation at 37 °C, cells were stained with crystal violet to visualize virus plaques. The results were expressed as the percentage of the number of virus plaques on enzyme-treated cells when compared to the number of virus plaques on mock-treated control cells.

Construction of HSV-1 gC mutants

Positively charged residues, most notably arginine and lysine, are crucial for the interaction of proteins with HS (Spillmann & Lindahl, 1994; Gallagher, 1995). This was the rationale for site-directed mutagenesis in which altered tandemly positioned positively charged amino acids within the N-terminal domain of HSV-1 gC were mutated to neutrally charged alanine residues in order to clarify their respective importance for HS binding. To study the local structure within a previously suggested HS-binding domain (Trybala et al., 1994), the hydrophobic amino acids Val140 and Ile142 were mutated to a non-hydrophobic residue (threonine), and Asn148 was mutated to alanine to abolish the glycosylation site at this position (Olofsson et al., 1999). In Fig. 1, the N-terminal region of HSV-1 gC is shown and the specific residues that were altered in each mutant virus are indicated. Mutant strains are designated by the use of the one-letter code for altered amino acids, e.g. R143,145A indicates that the arginine residues at positions 143 and 145 were replaced with alanine residues.

A total of 13 gC mutants and a gC-rescued variant of gC^–39 were constructed. Purified clones were subjected to Southern blot analysis to verify successful recombination (i.e. at the same genomic position as in the wild-type HSV-1 gC-positive strain) between the mutated gC gene and the gC^–39 strain DNA. All mutant viruses were then sequenced (amino acids 11–322 of the gC gene) to confirm the presence of the designed mutations. Sequence data for the virus mutants were as expected for all strains, except for mutant K(89,95)A+, which in addition to the designed mutations displayed an altered sequence between residues 75 and 80 (i.e. the sequence ⁷⁵KTTPTE⁸⁰ was unintentionally replaced with ⁷⁵NPRAHL⁸⁰), probably caused by the primer annealing to two separate but repetitive sequences in this region of the gC gene.

Expression of gC and the effect of mutations on the antigenic structure

gC, immunoprecipitated from all but two mutant strains, showed similar mobility in SDS–PAGE as gC from the KOS321 strain, indicating that the introduced mutations had not caused any major alterations in the size of gC. Two gC mutants, K(89,95)A+ and N148A, showed slightly higher mobility due to the loss of one N-glycosylation site in each mutant (Olofsson et al., 1999). Similar patterns were observed when the gC proteins, purified from the respective virus mutant by immunoaffinity chromatography, were subjected to SDS–PAGE (Fig. 2). Coomassie blue staining of gC from mutants R(151,155)A and R(155,160)A was weaker because a lesser amount of protein was electrophoresed.

Fig. 2. Electrophoretic analysis of gC purified from all constructed viruses as well as from the KOS321 strain. Infected GMK-AH1 cells and pelleted EX virus were lysed and the glycoproteins were clarified by centrifugation. The gC variants were bound onto an immunosorbent column, using MAb C4H11B6 to adsorb gC, and eluted with a low pH buffer. Equal volumes of the different preparations of purified gC were run on a 10 % SDS gel and stained with Coomassie blue.

To investigate the extent of conformational changes introduced into gC by the respective mutations, reactivity with a panel of anti-gC MAbs was assayed (Table 1). MAbs C1, C2 and C11, all of which are reactive with antigenic site I encompassing at least residues 307–373 at the C-terminal domain of gC (Wu et al., 1990), bound to all HSV-1 gC mutants, indicating that the introduced sequence mutations did not affect the structure of this antigenic site. However, MAbs C3, C8, C10, C13 and B1C1, which are all reactive with antigenic site II, previously delimited to residues 129–247 (Wu et al., 1990), gave different patterns of reactivity with gC in some mutants. MAbs C8, C10, C13 and B1C1 did not recognize gC produced by mutants R(143,145)A, R(145,147)A, R(151,155)A and R(155,160)A, with the exception of MAb B1C1, which showed some reactivity to the R(155,160)A mutant. MAb C3 was unable to bind to the R(129,130)A mutant or to the I142T mutant and was impaired in reactivity with the R(151,155)A and R(155,160)A mutants. These results were in agreement with the earlier mapping of these MAbs determined by sequencing the respective MAb-resistant (mar) mutants (Wu et al., 1990; Trybala et al., 1994), with one additional structural information provided: MAb C3, which was found previously to rely on Glu176 for binding, was also dependent on some or all of the residues Arg129, Arg130, Ile142, Arg151, Arg155 and Arg160 for efficient binding. MAb C2H12 (preliminarily mapped by pepscan to ²⁰¹PHVLW, unpublished observation) and MAb C4H11B6 (unmapped) recognized all of the HSV-1 mutant gC proteins. When used to quantify gC on purified viruses, however, MAb C4H11B6 gave similar absorbance values for all of the mutants at the different virus concentrations tested. Furthermore, comparisons of the amounts of virus needed to induce similar absorbance values after incubation with the gC and the gE MAbs, respectively, gave identical ratios for all mutants tested. Altogether, these results indicated that the induced mutations caused limited alterations within antigenic site II of HSV-1 gC, without affecting the incorporation of gC to the virus or the antigenic structure of site I.

Table 1. Binding of gC-reactive MAbs to GMK-AH1 cells infected with HSV-1 gC mutant viruses

MAb binding to HSV-1 gC mutant virus-infected cells was compared with binding to cells infected with the wild-type strain KOS321 and is graded as full (+), similar to that found on KOS321-infected cells, impaired (decreased) (±) or none (–).

* Data regarding epitope localization of the gC MAbs are based on sequencing of mar mutant strains (Wu et al., 1990; Trybala et al., 1994) except for the epitope of C2H12, which was preliminary mapped by pepscan.

ND, Not done.

Binding of purified mutant gC to isolated HS chains

To evaluate the direct interaction between gC and HS in solution, gC, purified from the mutated viruses and strain KOS321, was incubated with radiolabelled HS isolated from HEp-2 and GMK-AH1 cells. The gC–HS complexes were trapped on nitrocellulose filters and quantified (Maccarana et al., 1993). gC purified from strain KOS321 adsorbed approximately 25 % of input HS derived from HEp-2 cells and 15 % of HS from GMK-AH1 cells. Material from cells infected with the gC^– HSV-1 strain gC^–39 were subjected to mock immunoaffinity purification to serve as a negative control. In the binding assay, this preparation trapped less than 3 % of the labelled HS from both cell types, which was comparable to the background controls.

gC preparations purified from mutants R(143,145)A and R(151,155)A were found to be profoundly impaired in binding HS derived from HEp-2 cells (Fig. 3). Their binding activities were comparable to those of the negative control. Furthermore, gC proteins derived from the R(129,130)A, I142A, R(145,147)A and R(155,160)A mutants were also shown to poorly interact with HS from both GMK-AH1 and HEp-2 cell types. Intermediate-to-efficient binders of HS were gC proteins mutated at residues Lys114 and Arg117, Arg117 and Lys120, and Arg135 and Arg139, while gC from the remaining mutants was unimpaired in HS-binding. Similar data were found with HS derived from GMK-AH1 cells (data not shown). These results indicated that the cationic and hydrophobic amino acid residues localized at the base of the Cys127–Cys144 loop and close to the C-terminal region thereof were essential for HS binding.

Fig. 3. Binding of HSV-1 gC to HS isolated from HEp-2 cells. Preparations of gC (0.5 µg) purified from the respective HSV-1 gC mutants were incubated in PBS–BSA with 3000 c.p.m. of ³⁵S-labelled HS for 90 min at room temperature. The gC–HS complexes formed were trapped on a nitrocellulose filter and the amount of labelled HS was determined from two separate experiments.

HSV-1 gC mutants deficient in HS binding showed impaired HA

In contrast to myeloid cells, erythopoietic cells express HS during differentiation (Drzeniek et al., 1999). Earlier work has shown that HSV-1-induced HA of murine erythrocytes relies on the binding of gC, but not of other envelope glycoproteins, to HS-like molecules on the red blood cell surface (Trybala et al., 1993). The HA abilities of HSV-1 gC mutants are shown in Table 2. The results are expressed as the highest dilution of virus antigen that induced HA, and the type of HA, with regard to EX virus, is designated as strong/complete (++), weak/incomplete (+) or, if the antigen is unable to induce HA, none (–). EX virus preparations of mutants I142T, R(129,130)A, R(143,145)A, R(145,147)A, R(151,155)A and R(155,160)A showed no HA activity. Mutants displaying intermediate impairment (denoted in Table 2 as '+' and giving HA titres of between >2 and <256) were (K114,R117)A, (R117,K120)A and R(135,139)A. In contrast, HSV-1 gC mutants K(89,95)A+, K(105,107)A, V140T and N148A demonstrated no impairment in HA activity.

Table 2. HA of HSV-1 gC mutant viruses

* CA and EX HA antigens were prepared from virus-infected GMK-AH1 cells. Results are expressed as the highest dilution that induced haemagglutination.

† The type of haemagglutination (regarding EX virus) is graded as complete (++), incomplete (+) or none (–).

Reduction of cell surface HS and assessment of gC function in attachment

Direct attachment assays of purified radiolabelled virus with HEp-2 and Vero cells require high doses of virus to saturate the abundant cellular receptors (HS) and to show differences between a gC-positive and a gC-negative HSV-1 strain due to the remaining binding activity of gB (Herold et al., 1991). Since reduction of the amount of HS increased the virus dependency on gC, we determined the optimal reduction of cell surface HS for studies of gC-mediated attachment. To this end, HEp-2 monolayers were pre-treated with heparinase (0.08–8 units/ml) before KOS321 or gC^–39 strains were allowed to adsorb to the cells at 4 °C for 1 h. The c.p.m. ratio for bound:added KOS321 virus was 3318:9480 (35 %) and the corresponding c.p.m. ratio for gC^–39 was 1549:7041 (22 %). The percentage of radiolabelled virus that bound to the pre-treated cells as compared to mock-treated controls is shown in Fig. 4. The most pronounced difference between KOS321 and gC^–39 binding was achieved after pre-treatment with 0.25 units/ml of heparinase. This enzyme concentration was thus selected for subsequent attachment studies.

Fig. 4. Attachment of KOS321 and gC^–39 strains to HEp-2 cells pre-treated with heparinase. Duplicates of HEp-2 cell monolayers were pre-treated with serial twofold dilutions of heparinase, 8–0.008 units/ml, for 1 h at 37 °C and thereafter blocked with 1 % BSA and left to adsorb ³H-labelled virus for 1 h at 4 °C. The results are expressed as the percentage of attached labelled virus particles compared to control cells untreated with heparinase. The arrow indicates the heparinase concentration at which the binding properties of the two viruses differed the most. This pre-treatment was then selected for the subsequent attachment assays of the gC-1 mutants (see Fig. 5).

Mutations of HSV-1 gC that reduced virus attachment to cells pre-treated with low doses of heparinase

The attachment of HSV-1 gC mutants as well as the KOS321 and gC^–39 strains to heparinase-treated HEp-2 cells, in comparison to untreated cells, is shown in Fig. 5. The c.p.m. ratio for bound:added KOS321 virus was 12650:34700 (36 %) and the corresponding c.p.m. ratio for gC^–39 was 6700:27900 (24 %). The virus mutants could be divided into three groups as regards the efficiency of their binding to heparinase-treated cells: (i) unimpaired attachment was exhibited by the HSV-1 gC mutants K(89,95)A+, K(105,107)A, (K114,R117)A and (R117,K120)A, all with basic residues altered, as well as the V140T and N148A mutants; (ii) moderately impaired attachment was found for the R(129,130)A, R(135,139)A and I142T mutants; and (iii) profoundly impaired attachment to heparinase-treated cells (even more pronounced than that of the gC^–39 strain) was documented for the R(143,145)A, R(145,147)A, R(151,155)A and R(155,160)A mutants. The results for GMK-AH1 cells were similar (data not shown) and indicated a crucial role for the arginine residues clustered between positions 143 and 160 in binding to HS sequences remaining on cell surfaces after low-dose heparinase treatment.

Fig. 5. Attachment of radiolabelled HSV-1 gC mutant viruses to HEp-2 cells pre-treated with heparinase. Duplicate cell monolayers were pre-treated with heparinase (0.25 units/ml) for 1 h at 37 °C and blocked for 1 h before virus was added and left for adsorption at 4 °C for 1 h. For each mutant, the amount of virus adsorbed to the heparinase-treated cells was determined and expressed as a percentage of control, i.e. the amount of the same virus adsorbed to untreated cells.

HSV-1 gC mutants deficient in HS binding showed reduced infectivity in heparinase-treated cells

To determine whether the observed impairment in attachment to cell surfaces with reduced amounts of HS might influence the infectivity of the different mutants studied, GMK-AH1 cells were treated with 1 unit/ml of heparinase before addition of virus. The results expressed as a percentage of the number of plaques observed in mock-treated cells are shown in Fig. 6. The HSV-1 gC-rescued mutant as well as the KOS321 strain showed almost no decrease in infectivity after pre-treatment of cells with this specific dose of heparinase. Mutated strains carrying substitutions for basic residues stretching from Lys75 to Lys120 as well as Arg135 and Arg139 exhibited little decrease in their infectivity. In contrast, the gC mutants R(129,130)A, I142T, R(143,145)A, R(145,147)A, R(151,155)A and R(155,160)A showed a profound decrease (more pronounced than that of gC^–39) in their ability to infect the enzyme-treated cells. Hence, our interpretation is that virus mutants carrying gC defective in HS binding and attachment to cells were also impaired in their ability to infect GMK-AH1 cells exhibiting reduced amount of cell surface HS.

Fig. 6. Infectivity of HSV-1 gC mutant viruses on heparinase-treated GMK-AH1 cells. Viruses were pre-diluted and approximately 200–300 p.f.u. of each mutant was used for comparison of infectivity on untreated and heparinase-treated cells. Monolayers of GMK-AH1 cells in 6-well plates were pre-treated with heparinase (1 unit/ml) for 1 h at 37 °C. Virus was then allowed to adsorb to the cells for 30 min at 37 °C, whereafter the cells were covered with methylcellulose and incubated in 37 °C to allow virus plaque formation. Cells were stained with crystal violet and plaques were counted. The results are expressed as the number of plaques formed on heparinase-treated cells as compared to the control, i.e. the number of plaques formed on monolayers not subjected to pre-treatment with heparinase.

Based on a scanning mutagenesis of virtually all positively charged residues as well as some hydrophobic residues within the N-terminal domain (amino acids 26–227) of mature HSV-1 gC, we propose that residues ¹⁴²IRCRFRNSTRMEFRLQIWR¹⁶⁰ together with ¹²⁹RR¹³⁰ (where bold letters indicate active amino acids) constitute a major HS-binding attachment domain of HSV-1 gC. Despite the fact that the HS-binding activity of this stretch depends on the cluster of arginines typical for proteins displaying this function, it does not fit into any hitherto proposed consensus heparin-binding sequence (Hileman et al., 1998). Most likely, some or all of these arginine residues participate in electrostatic interactions with the 6-O- and 2-O-sulfate groups in the dodecamer of the HS chain that was shown previously to represent a minimum sequence required for binding to gC (Feyzi et al., 1997).

Although a detailed structural characterization of HSV-1 gC is lacking, the disulfide bonds between the cysteine residues of the protein have been resolved (Rux et al., 1996). The residues within antigenic site II involved in HS binding are situated at or near a 16 residue loop (Cys127–Cys144), which is well-conserved among mammalian herpesviruses, as judged from reported DNA sequences (Fitzpatrick et al., 1989). In PrV, the HS-binding domain of gC was shown previously to be composed of three functionally redundant units located in the vicinity of the homologous loop region of that protein (Flynn & Ryan, 1996). The architecture of the HS-binding domain in HSV-1 gC shows a variation of a similar theme. In addition, other regions within the HSV-1 gC antigenic site II, such as Gly247 (Wu et al., 1990), selected during preliminary mapping of the HS-binding site on gC by the use of mar mutants (Trybala et al., 1994), may well participate in HS interaction, but this question was not addressed in the present work.

The HS-binding domain of gC, identified in the present work, has an amphipathic character and the hydrophobic residue Ile142, but not Val140, was found to be decisive for HS binding of the protein as well as important for virus attachment and infectivity. In addition, results from ongoing studies with a single mutant carrying a Phe146-to-Thr mutation suggest an importance of this hydrophobic residue (unpublished observation). Taken together, these findings indicate that the binding of gC to HS is not a purely electrostatic interaction and that nonionic forces may contribute to free energy of binding, as was reported previously to be the case for heparin–basic fibroblast growth factor interactions (Thompson et al., 1994).

The findings that gC binding to HS is difficult to saturate and that gB may replace gC function in HS-driven attachment to the cell surface (Herold et al., 1991, 1994; Gerber et al., 1995; Laquerre et al., 1998) pose a difficulty when analysing effects of minor structural alterations in gC on virus adsorption. Here, we screened restricted mutations of the gC gene in otherwise unaltered viruses in two attachment assays selected to enable a gB-independent evaluation of gC function. The results from the gC/HS-specific HSV-1 HA (Trybala et al., 1993) showed that those mutants from which purified gC proteins were found to be defective in HS interaction were also impaired in binding to murine erythrocytes. Furthermore, low-dose heparinase treatment provided the means to study impairment of gC function during attachment to living cells, probably explained by fewer HS molecules being available for binding. Again, results were in agreement with those from the HS-binding abilities of gC proteins, with one exception: the R(135,139)A mutant, only slightly impaired in HS binding and HA activity, showed a more pronounced disability in attachment to heparinase-treated cells, suggesting that these residues, situated at the apex of the loop, may be important for an additional interaction during the attachment step. However, the only moderately reduced infectivity of this mutant suggests that the results should be interpreted with caution, considering the methodological differences that exist between these experimental assays.

In the complex early events of -herpesvirus infection, a dispensable role for gC for overall infection of cultured cells was shown for HSV, PrV and BHV-1 based on whole gene deletion mutants (reviewed by Spear, 1993). In line with this, even the gC mutants that were almost devoid of HS binding, such as R(143,145)A, replicated well in cell cultures in the present study. However, when infectivity was assayed after low-dose heparinase treatment of cells, this mutant appeared to be even less infective than the gC^– virus. This finding indicates that the gC–HS interaction is also important for virus infectivity and that the presence of gC molecules impaired in HS binding might hinder the redundancy of gB, at least in a milieu where the number and/or length of HS chains is reduced. Under such conditions, a defect in the gC–HS interaction might constitute a bottle-neck that reduces virus infectivity.

By using HS-specific antibodies, the in vivo expression of specific HS epitopes was shown to be limited and tissue-specific (van Kuppevelt et al., 1998). The ability of gC to fish out a subset of HS molecules (Feyzi et al., 1997) suggested a specificity of this interaction to which both molecules contributed. The understanding of the biosynthesis of HS, of specific importance for HSV infectivity, was increased by the characterization of the EXT-1 gene, which is one of a family of HS-specific glycosyltransferases (Lind et al., 1998; McCormick et al., 1998). In addition, the report that later stages of HSV-1 entry mediated by gD binding to a family of immunoglobulin-like cell surface molecules (Cocchi et al., 1998; Geraghty et al., 1998) can be fully substituted by gD binding to HS chains modified by a specific isoform of 3-O-sulfotransferase (Shukla et al., 1999) poses the question of a possible functional relationship between gD/gC and cell surface HS molecules. Such studies might benefit from the availability of virus mutants defective in HS interaction, but which carry only point mutations in gC.

We thank Maria Johansson for skilful technical assistance. This work was supported by the Swedish Medical Research Council (grant no. 11225), the Sahlgren's University Hospital LUA foundation and the program 'Glycoconjugates in Biological Systems' sponsored by the Swedish Foundation for Strategic Research.

Bergström, T., Sjögren-Jansson, E., Jeansson, S. & Lycke, E. (1992). Mapping neuroinvasiveness of the herpes simplex virus type 1 encephalitis-inducing strain 2762 by the use of monoclonal antibodies. Molecular and Cellular Probes 6, 41–49.

Cocchi, F., Menotti, L., Mirandola, P., Lopez, M. & Campadelli-Fiume, G. (1998). The ectodomain of a novel member of the immunoglobulin subfamily related to the poliovirus receptor has the attributes of a bona fide receptor for herpes simplex virus types 1 and 2 in human cells. Journal of Virology 72, 9992–10002.

Drzeniek, Z., Stöcker, G., Siebertz, B., Just, U., Schroeder, T., Ostertag, W. & Haubeck, H.-D. (1999). Heparan sulfate proteoglycan expression is induced during early erythroid differentiation of multipotent hematopoetic stem cells. Blood 93, 2884–2897.

Feyzi, E., Trybala, E., Bergström, T., Lindahl, U. & Spillmann, D. (1997). Structural requirement of heparan sulfate for interaction with herpes simplex virus type 1 virions and isolated glycoprotein C. Journal of Biological Chemistry 272, 24850–24857.

Fitzpatrick, D. R., Babiuk, L. A. & Zamb, T. J. (1989). Nucleotide sequence of bovine herpesvirus type 1 glycoprotein gIII, a structural model for gIII as a new member of the immunoglobulin superfamily, and implications for the homologous glycoproteins of other herpesviruses. Virology 173, 46–57.

Flynn, S. J. & Ryan, P. (1996). The receptor-binding domain of pseudorabies virus glycoprotein gC is composed of multiple discrete units that are functionally redundant. Journal of Virology 70, 1355–1364.

Fuller, A. O. & Spear, P. G. (1985). Specificities of monoclonal and polyclonal antibodies that inhibit adsorption of herpes simplex virus to cells and lack of inhibition by potent neutralizing antibodies. Journal of Virology 55, 475–482.

Gallagher, J. T. (1995). Heparan sulphate and protein recognition. Binding specificities and activation mechanisms. Advances in Experimental Medicine and Biology 376, 125–134.

Geraghty, R. J., Krummenacher, C., Cohen, G. H., Eisenberg, R. J. & Spear, P. G. (1998). Entry of alphaherpesviruses mediated by poliovirus receptor-related protein 1 and poliovirus receptor. Science 280, 1618–1620.

Gerber, S. I., Beval, B. J. & Herold, B. C. (1995). Differences in the role of glycoprotein C of HSV-1 and HSV-2 in viral binding may contribute to serotype differences in cell tropism. Virology 214, 29–39.

Graham, F. L. & van der Eb, A. J. (1973). A new technique for the assay of infectivity of human adenovirus 5 DNA. Virology 52, 456–467.

Herold, B. C., WuDunn, D., Soltys, N. & Spear, P. G. (1991). Glycoprotein C of herpes simplex virus type 1 plays a principal role in the adsorption of virus to cells and in infectivity. Journal of Virology 65, 1090–1098.

Herold, B. C., Visalli, R. J., Susmarski, N., Brandt, C. R. & Spear, P. G. (1994). Glycoprotein C-independent binding of herpes simplex virus to cells requires cell surface heparan sulphate and glycoprotein B. Journal of General Virology 75, 1211–1222.

Hileman, R. E., Fromm, J. R., Weiler, J. M. & Linhardt, R. J. (1998). Glycosaminoglycan–protein interactions: definition of consensus sites in glycosaminoglycan binding proteins. Bioessays 20, 156–167.

Holland, T. C., Marlin, S. D., Levine, M. & Glorioso, J. (1983). Antigenic variants of herpes simplex virus selected with glycoprotein-specific monoclonal antibodies. Journal of Virology 45, 672–682.

Holland, T. C., Homa, F. L., Marlin, S. D., Levine, M. & Glorioso, J. (1984). Herpes simplex virus type 1 glycoprotein C-negative mutants exhibit multiple phenotypes, including secretion of truncated glycoproteins. Journal of Virology 52, 566–574.

Homa, F. L., Purifoy, D. J. M., Glorioso, J. C. & Levine, M. (1986). Molecular basis of the glycoprotein C-negative phenotypes of herpes simplex virus type 1 mutants selected with a virus-neutralizing monoclonal antibody. Journal of Virology 58, 281–289.

Immergluck, L. C., Domowicz, M. S., Schwartz, N. B. & Herold, B. C. (1998). Viral and cellular requirements for entry of herpes simplex virus type 1 into primary neuronal cells. Journal of General Virology 79, 549–559.

Karger, A., Saalmuller, A., Tufaro, F., Banfield, B. W. & Mettenleiter, T. C. (1995). Cell surface proteoglycans are not essential for infection by pseudorabies virus. Journal of Virology 69, 3482–3489.

Laquerre, S., Argnani, R., Anderson, D. B., Zucchini, S., Manservigi, R. & Glorioso, J. C. (1998). Heparan sulfate proteoglycan binding by herpes simplex virus type 1 glycoproteins B and C, which differ in their contributions to virus attachment, penetration, and cell-to-cell spread. Journal of Virology 72, 6119–6130.

Lind, T., Tufaro, F., McCormick, C., Lindahl, U. & Lidholt, K. (1998). The putative tumor suppressors EXT1 and EXT2 are glycosyltransferases required for the biosynthesis of heparan sulfate. Journal of Biological Chemistry 273, 26265–26268.

Lyon, M., Deakin, J. A., Mizuno, K., Nakamura, T. & Gallagher, J. T. (1994). Interaction of hepatocyte growth factor with heparan sulfate. Elucidation of the major heparan sulfate structural determinants. Journal of Biological Chemistry 269, 11216–11223.

Maccarana, M., Casu, B. & Lindahl, U. (1993). Minimal sequence of heparin/heparan sulfate required for binding of basic fibroblast growth factor. Journal of Biological Chemistry 268, 23898–23905.

McCormick, C., Leduc, Y., Martindale, D., Mattison, K., Esford, L. E., Dyer, A. P. & Tufaro, F. (1998). The putative tumour suppressor EXT1 alters the expression of cell-surface heparan sulfate. Nature Genetics 19, 158–161.

Maniatis, T. E., Fritsch, E. F. & Sambrook, J. (1982). Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Marlin, S. D., Holland, T. C., Levine, M. & Glorioso, J. C. (1985). Epitopes of herpes simplex virus type 1 glycoprotein gC are clustered in two distinct antigenic sites. Journal of Virology 53, 128–136.

Mettenleiter, T. C., Zsak, L., Zuckermann, F., Sugg, N., Kern, H. & Ben-Porat, T. (1990). Interaction of glycoprotein gIII with a cellular heparinlike substance mediates adsoption of pseudorabies virus. Journal of Virology 64, 278–286.

Okazaki, K., Mitsuzaki, T., Sugahara, Y., Okada, J., Hasebe, M., Iwamura, Y., Ohnishi, M., Kanno, T., Shimizu, M., Honda, E. & Kono, Y. (1991). BHV-1 adsorption is mediated by the interaction of glycoprotein gIII with heparin-like moiety on the cell surface. Virology 181, 666–670.

Olofsson, S., Bolmstedt, A., Biller, M., Mårdberg, K., Leckner, J., Malmström, B. G., Trybala, E. & Bergström, T. (1999). The role of a single N-linked glycosylation site for a functional epitope of herpes simplex virus type 1 envelope glycoprotein gC. Glycobiology 9, 73–81.

Rux, A. H., Moore, W. T., Lambris, J. D., Abrams, W. R., Peng, C., Friedman, H. M., Cohen, G. H. & Eisenberg, R. J. (1996). Disulfide bond structure determination and biochemical analysis of glycoprotein C from herpes simplex virus. Journal of Virology 70, 5455–5465.

Salmivirta, M., Lidholt, K. & Lindahl, U. (1996). Heparan sulfate: a piece of information. FASEB Journal 10, 1270–1279.

Shukla, D., Liu, J., Blaiklock, P., Shworak, N. W., Bai, X., Esko, J. D., Cohen, G. H., Eisenberg, R. J., Rosenberg, R. D. & Spear, P. G. (1999). A novel role for 3-O-sulfated heparan sulfate in herpes simplex 1 entry. Cell 99, 13–22.

Spear, P. G. (1993). Entry of alphaherpesviruses into cells. Seminars in Virology 4, 167–180.

Spillmann, D. & Lindahl, U. (1994). Glycosaminoglycan–protein interactions: a question of specificity. Current Opinion in Structural Biology 4, 677–682.

Svennerholm, B., Jeansson, S., Vahlne, A. & Lycke, E. (1991). Involvement of glycoprotein C (gC) in adsorption of herpes simplex virus type 1 (HSV-1) to the cell. Archives of Virology 120, 273–279.

Tal-Singer, R., Peng, C., Ponce de Leon, M., Abrams, W. R., Banfield, B. W., Tufaro, F., Cohen, G. H. & Eisenberg, R. J. (1995). Interaction of herpes simplex virus glycoprotein gC with mammalian cell surface molecules. Journal of Virology 69, 4471–4483.

Thompson, L. D., Pantoliano, M. W. & Springer, B. A. (1994). Energetic characterization of the basic fibroblast growth factor–heparin interaction: identification of the heparin binding domain. Biochemistry 33, 3831–3840.

Trybala, E., Svennerholm, B., Bergström, T., Olofsson, S., Jeansson, S. & Goodman, J. L. (1993). Herpes simplex virus type 1-induced hemagglutination: glycoprotein C mediates virus binding to erythrocyte surface heparan sulfate. Journal of Virology 67, 1278–1285.

Trybala, E., Bergström, T., Svennerholm, B., Jeansson, S., Glorioso, J. C. & Olofsson, S. (1994). Localization of a functional site of herpes simplex virus type 1 glycoprotein C involved in binding to cell surface heparan sulphate. Journal of General Virology 75, 743–752.

Trybala, E., Bergström, T., Spillmann, D., Svennerholm, B., Flynn, S. & Ryan, P. (1998). Interaction between pseudorabies virus and heparin/heparan sulfate. Pseudorabies virus mutants differ in their interaction with heparin/heparan sulfate when altered for specific glycoprotein C heparin-binding domain. Journal of Biological Chemistry 273, 5047–5052.

van Kuppevelt, T., Dennissen, H. M., van Venrooij, W. J., Hoet, R. & Veerkamp, J. H. (1998). Generation and application of type-specific anti-heparan sulfate antibodies using phage display technology. Journal of Biological Chemistry 273, 12960–12966.

Wu, C.-T. B., Levine, M., Homa, F., Highlander, S. L. & Glorioso, J. C. (1990). Characterization of the antigenic structure of herpes simplex virus type 1 glycoprotein C through DNA sequence analysis of monoclonal antibody-resistant mutants. Journal of Virology 64, 856–863.

WuDunn, D. & Spear, P. G. (1989). Initial interaction of herpes simplex virus with cells is binding to heparan sulfate. Journal of Virology 63, 52–58.

Zhu, Z., Gerson, M. D., Ambron, R., Gabel, C. H. & Gerson, A. A. (1995). Infection of cells by varicella zoster virus: inhibition of viral entry by mannose 6-phosphate and heparin. Proceedings of the National Academy of Sciences, USA 92, 3546–3550.

JGV Direct table of contents

© 2001 SGM

This article is now available in the August 2001 print issue of JGV (vol. 82, 1941–1950). The complete issue of the journal may be seen in electronic form on JGV Online.